• Careers
  • Contact Us
  • Blog
  • REQUEST A DEMO

MedHOK Strategic Insights Blog

Best Practice Memo on Encounter Data Submission is Telling Sign That CMS Has Not Given Up On Risk Adjustment and Payment Accuracy Changes

Marc Ryan | August 10, 2017

u need 2 know.jpgOn June 22, 2017, the Centers for Medicare and Medicaid Services (CMS) issued a best practice memo related to encounter data submission in the Medicare Advantage (MA) program. To us it signals that, despite some missteps and retreats along the way, CMS will bring their encounter data submission vision to sunset RAPS and fully risk adjust using encounters to fruition in the not-too-distant future.

Risk adjustment transformation stalled

MA has used a proprietary Risk Adjustment Processing System (RAPS) format and process that was relatively easy for plans to compile, submit, and remediate if records were rejected by CMS. In 2012, CMS began running the Encounter Data Processing System (EDPS or 837) program in parallel with RAPS. By 2015, CMS began using 837 submissions to supplement data for risk scores submitted through RAPS. In 2016, the first blending occurred, with 90 percent of risk adjustment factored using RAPS and 10 percent using EDPS.

It has been a rocky road since then. First, CMS had their own issues related to the acceptance and processing of 837 files and records. Second, and more importantly, many plans were simply not ready and are still poorly placed to succeed using the new process. Unless you were a plan that also had robust Medicaid 837 encounter submission in a given state or states, plans faced (and still do) huge revenue impacts because of it. So much so that while in 2017 risk adjustment is weighted 75 percent RAPS and 25 percent EDPS (and was scheduled to go to 50-50 in 2018 and have EDPS account for 100 percent by 2020), in the 2018 Call Letter, CMS announced it was forestalling further implementation. In fact, due to the issues (perhaps again on both sides), the 2018 blend goes back a little to 85 percent RAPS and 15 percent EDPS. A life-saver financially for many plans.

History shows, CMS is usually committed to changes, even if barriers exist

As we note above, the June 22, 2017 Health Plan Management System (HPMS) memo is a signal to plans to get their resources and processes in place. Plans should not rest even with the CMS respite. As we know, especially with CMS, history repeats itself (or perhaps more accurately to paraphrase Mark Twain, it rhymes). CMS initially stumbled on putting in place risk adjustment as a comprehensive rate-setting system years ago. But it worked out the issues with plans and went ahead. The same happened with the Star program over the years. The roll out of today’s rigorous compliance and audit infrastructure saw similar issues. In all of these cases, CMS accommodated along the way and issued best practice memos. And all of these three programs today represent how government can be run smoothly and in the best interest of citizens. So plans are likely deluding themselves if they don’t think CMS will ultimately push forward with sunsetting RAPS and fully risk adjusting using EDPS. As we explain later in this blog, CMS wants this migration to occur because they see accurate reimbursement based on sophisticated encounters as a way to grapple with and rein in the rampant fraud, waste and abuse (FWA) problem seen in government healthcare programs. Encounters don’t lie.

The complexity of EDPS compared with RAPS

Setting up the right EDPS infrastructure is not something done overnight. It is painstaking, arduous, and labyrinthine. The reason: RAPS is a simple straightforward format with a limited number of fields. A RAPS file is easily submitted. If a record on the file fails, it is easily remediated because of the simple format. But with 837 processing, obstacles are much greater. An 837 record has the complete history of an encounter or claim in a sophisticated loop structure. 837s can fail at the file and record level multiple times and for various reasons. Establishing root cause takes time and remediation is time-consuming. Thus, so-called “leakage” (the inability to identify the reasons for encounter rejection and losing the opportunity to qualify an encounter) is much greater under EDPS than RAPS. Most important, plans have lived or died on adding to their revenue by running supplemental RAPS processes for diagnoses not previously reported on encounter or claim submissions. This is usually done through a chart review, with RAPS records created swiftly. While this is possible, and practiced, under EDPS it is much more complex and time-consuming as every record has to go back through the full EDPS process.

  Issue

  RAPS

  EDPS

  Data/format

  5 data elements

  Utilizes all elements of a HIPAA
  standard 5010 format 837

  Service/POS

  Limited

  All

  Submission

  Relatively straightforward
  submission

  Sophisticated encounter
  submission process

  Pathway

  FERAS level and record level

  Encounter gateway for file and
  records

  Leakage

  Some, but can recover

  Significant and hard to track

  Analysis and Remediation

  Relatively simple

  Much more complex

  Supplemental

  Major; straightforward

  More minor; complete

 

How plans can excel at EDPS – MedHOK’s vision

How can plans weather the storm and successfully adapt to the change from RAPS to EDPS? From our standpoint, here are some general best practices plans should employ to safeguard revenue:

  • It is imperative that plans implement an end-to-end, thorough encounter data gathering and submission process. This touches many different departments of a health plan, including provider relations, IT, clinical, and finance. It also starts with education of the provider and goes throughout the submission and remediation process. This includes overseeing capitated providers, FFS providers, and downstream subcontractors for their accurate and timely encounter submission.
  • Additionally, plans will need to ensure that all claims and encounters move seamlessly, cleanly, and swiftly through the claims and other systems and that compliant 837 data is created for submission to CMS on cycle. Many plans still struggle with this. MA plans have 13 months for EDI submissions. Practically speaking, MA plans will need to build aggressive timeframes to submit and then resubmit encounters as necessary to ensure all diagnoses are properly credited, remediation activities can occur timely, and timeframes don’t get stale and lead providers and plans to move on to other things (we know providers will look way back in time to get paid but they are unlikely to put in the administrative effort if there is little or no tangible benefit to them). Again, all this takes collaboration across the health plan, including provider relations, information technology, finance and more.
  • Find ways to build incentives, within the bounds of the regulations, to ensure providers cooperate in the endeavor. Incentivize timely, accurate and complete submission by providers. Plans often see issues with capitated primary care providers. They should submit encounters for services but sometimes do not as they don’t get paid by claim. These submissions are essential as these are where most diagnoses come from.
  • Capture both RAPS Return Files and Encounter Data Return Files during the existing transition process. Conduct remediation of both RAPS and EDPS submissions. This is especially key for EDPS, where files can be rejected as well as individual records for multiple times and reasons.
  • Conduct robust financial reconciliation on both RAPS an EDPS submissions to determine where deltas exist between the two formats/submissions. Thoroughly look at differences between the RAPS Return Files and the various EDPS reports (largely MAO-002 and MAO-004). Examine the quality (accuracy and timeliness), service utilization (completeness and gaps), and financial (where are the financial differences and why) aspects between the submissions. For both RAPS and EDPS, compare frequency/count by diagnosis in the aggregate. For both RAPS and EDPS, identify the number of unique members with one or more diagnoses. Compare HCCs that are accepted in RAPS and not accepted in EDPS. Log error reasons and frequency. Analyze both RAPS and EDPS to ascertain “falloffs” (when chronic conditions are not reported at least once per year) and “opportunities” (where higher reimbursement may exist within a disease cluster or a new qualify cluster may exist). Inevitably plans will find areas where EDPS lags RAPS. This could be by provider, by diagnosis, between institutional and professional submissions, and even perhaps by state/region. CMS just released its third iteration of MAO-004 (there are still some issues, including with regard to DME) that should provide greater insight. By adding additional fields, the latest MAO-004 shows both submissions that passed and did not pass and insight into the filtering  logic.

It is important to note that plans are a bit on their own to analyze RAPS and EDPS differences. In payment year 2017, CMS had planned to introduce the blend of RAPS (75%) and EDPS (25%) for midyear risk scores. They backed off and will maintain 100% RAPS, with reconciliation against the blend in mid-2018. This would have provided insight into differences for each plan. Therefore, following these best practices is now even more important. 

How plans can excel at EDPS – CMS suggestions

In its June 22, 2017, memo, CMS also offered more specific best practices it has complied since inception of the parallel process of encounter submission. Some of these overlap with our recommendations above, but we have included them again to underscore the significance.

  • Reconcile MAO-002 and MAO-004 reports. CMS notes that the 277CA and MAO-0021 reports provide record- and line-level disposition status and associated edit codes that will help plans understand what was and was not accepted and why. This data can be used both for rejects and for informational edits. Plans should keep a data warehouse of these records and mine this data to determine major areas of errors on submission.
  • CMS encourages plans to review submission report cards regularly. CMS began issuing report cards in September 2015. These report cards give valuable information from CMS on its assessment of plan-specific submission volume, submission frequency, and accuracy of encounters. CMS notes that low submissions may be an indication that there are gaps in submission.
  • As we proposed above, CMS believes that an internal-external, inter-disciplinary encounter task force is vital given the fact that encounters touch so many areas. Composition should span the departments of the plan and include providers and subcontractors that collect encounters.
  • Also as we note, CMS says provider education with feedback loops is key as they may not understand the significance of encounters in rate-setting and, ultimately, their reimbursement. They, too, note the issue we outlined above on capitated providers.
  • CMS also encourages comparison of RAPS and EDPS to understand completeness and accuracy. CMS recommends performing analysis on encounter data submissions to ascertain variation in encounter data volume (weekly and monthly patterns and anomalies).
  • Prior to submitting chart review records, CMS advises plans to check the Internal Control Number (ICN), header date of service, and diagnosis code(s) on the chart review against all other chart review submissions and encounter data record submissions to ensure that duplicate diagnosis information is not submitted.
  • CMS indicates that plans should use the Encounter Data Front End System (EDFES) test environment to ensure the correct submission practice. If a plan is unclear on how to properly code specific scenarios, it may send a test encounter data record to the EDFES test environment prior to official submission. This will help understanding of how edits are applied in the system.
  • CMS says plans should verify that their internal processes complete. Plans may have a number of systems that must complete steps before a successful submission to CMS.
  • CMS encourages reduction of provider paper claim submissions to reduce error and administrative burden.

Why is all this important?  There are major revenue implications
There are huge implications for revenue for plans. In a January 2017 white paper published by Milliman, the actuary firm studied differences in EDPS and RAPS risk scores for about 900,000 lives across 45 plans. The firm concluded that EDPS rendered a median difference of 4 percent lower Part C revenue (for Special Needs Plans (SNPs) and mainstream MA plans combined). While 87 percent of scores were the same for members, 13% of members had different scores, with 90% of the differences being lower. SNPs alone suffered an over 5 percent revenue reduction.

The revenue impact is hugely significant. While currently EDPS is a small percentage of the blend and represents a more modest financial revenue impact, over time the lower revenue would wipe out many plans' total margins if not tackled. And there is reason to believe the impact as it stands now is greater. While some plans refused to participate in the Milliman study for proprietary reasons or because they thought they had a good process in place, many more likely could not participate because they could not even supply the data because their EDPS processes are grossly immature.

Why is all this important? It is all about FWA

We know that FWA constitutes as much as 10% of overall national healthcare expenditures. Over 90% of FWA is overpayments of some sort.  And we are not talking just traditional providers in Medicare, but to the growing MA plan segment as well.

For 2015, the federal government estimates that as much as 9.5% of payments to MA plans represent improper payments. That is an estimated $14.1 billion on payments of $148.6 billion. Further, the federal government estimates that Medicare Advantage is the fourth-largest program for overpayments, behind traditional Medicare (FFS), Medicaid, and the Earned Income Tax Credit (EITC). These four programs are the only ones in the $10 billion plus category.

FFY 2015 Estimated Improper Payments in Federal Healthcare Programs

  Program

  Payments

  Improper
  Amount

  Percentage

  Medicare Fee For Service

  $358.3B

  $43.3B

  12.10%

  Medicaid

  $297.7B

  $29.1B

  9.80%

  Medicare Advantage Part C

  $148.6B

  $14.1B

  9.50%

  Medicare Part D

  $62.0B

  $2.2B

  3.60%

  State Children’s Health Insurance
  Program (CHIP)

  $9.3B

  $0.6B

  6.80%

 

While the executive branch and CMS are concerned about the statistics above, the MA overpayment problem has also become a focus of Congress. And it will involve both Democrats and Republicans. Health plans had much to worry about recently when Judiciary Chairman (as well as former Chairman and still influential member of the Finance Committee) Chuck Grassley, R-IA, generally a staunch supporter of MA,, essentially declared war on MA plans when it came to overpayments. He sent a letter to CMS Administrator Seema Verma demanding answers on why a mere fraction of overpayments detailed in a Government Accountability Office (GAO) report were recouped. He challenged Verma to “aggressively use the tools at its (CMS’) disposal to ensure that it is efficiently identifying fraud and subsequently implementing timely and fair remedies.” In this case it surrounded a lack of documentation on risk adjustment payments (more below), but focused a light on overpayments generally and the FWA problem globally.

Where plans will need to focus

CMS sees the following as ways to reduce overpayments and FWA in general in MA.

  • Continually refine rate-setting to ensure accurate recognition of the relative risk of each member. As we have seen, they are doing this on the demographic component side (through the community segment split-out) as well as the clinical component side (through regular updates to both Part C and D HCC models).
  • Enhance monthly enrollment and payment attestation and reporting.
  • Implement sophisticated encounter processing and rate-setting to reduce what they view as the “over-coding” in the RAPS system (more on this later).
  • Implementing a Risk Adjustment Data Validation (RADV) audit procedure to ensure plans can validate the data that they are being paid on.

More specifically, MA plans will need to meticulously reconcile enrollment and payment in the following big areas:

  • Enrollment discrepancies globally (do I really have this member for the timeframe CMS paid me?)
  • Payment in various categories is accurate. For example, is the plan being paid correctly for institutional, ESRD and the now six broken-out community categories (full benefit dual aged, full benefit dual disabled, partial benefit dual aged, partial benefit dual disabled, non-dual aged and non-dual disabled)? The community change significantly increases complexity for plans in determining proper classification and payment and reporting to CMS. What is the plan doing to reconcile the factors above as well as a myriad of other demographic indicators that impact revenue (State and County Code, Hospice, Out of Area, Age, Gender, Plan/Contract, Medicaid Status, Medicare Savings Program Status, Low Income Subsidy Status, etc.)?
  • Clinical risk adjustment factor portions of the payment for both Part C (HCC) and Part D (RxHCC). Are the HCCs correct and can a plan validate these with diagnoses from claims and medical records (see more below)? While plans by and large practice global enrollment reconciliation, very few plans today are performing true payment reconciliation given its complexity.

We would also note that while proper reconciliation of the above will guard against fines for overpayment, it also affords an opportunity for plans to ensure they are being properly reimbursed by CMS. CMS likes to focus on overpayment, but plans will need to ensure it is a two-way street to help them guard already dwindling margins.

The coming Risk Adjustment Data Validation (RADV) focus

After a rather slow start on RADV, CMS has now begun a yearly audit process that, when fully implemented, could recoup billions from MA plans. In response to the congressional actions mandating a crackdown on improper payments, CMS began its RADV program with a pilot audit of 5 contracts for the 2007 payment year as well as a selection of 32 additional contracts for the 2007 payment year. As of the last update by CMS in late 2015, these audits found about $13.8M in overpayments on the audit sample and plans were still in the dispute and appeal process. The GAO, the auditing and investigations arm of the U.S. Congress, recently issued a report (referenced by Grassley above) criticizing CMS’ approach to overpayments, indicating it has made little progress to prevent overpayments thus far and spent over $117 million on administration since 2010 while recouping just $14 million. It noted all but two of the plans audited had overpayments and overpayments were three to four times more likely than underpayments.

CMS has long been concerned that MA plan risk scores are demonstrably higher than those in Medicare FFS and has taken a number of steps to address what it views as “over-scoring,” including adjusting the MA rates downward via a coding modification in the rate-setting process. CMS, too, continually adjusts the HCC models and has recently made adjustments in diagnosis clusters and severity in part as a result of over-scoring concerns (e.g., diabetes).

But, with the GAO report, and the shot across the bow by Grassley, CMS will now have to ramp up its RADV efforts. The RADV process is meant to address these inequities on a plan-by-plan basis. Plans have argued that they do not inflate scores but, unlike a FFS provider, have an incentive to accurately code members. Yet at the same time, plans have challenges substantiating the diagnoses that lead to the risk adjustment on the clinical side of payments. These problems come from two main areas:

  • Providers may submit encounters and claims to plans without adequately documenting pertinent disease states or conditions on medical records. In the 2007 audits, only about 60% of diagnoses could be confirmed on audit. CMS says about three quarters of over payments are tied to insufficient medical documentation.
  • Plans are unable to track down medical records from providers for submitted diagnoses.

How can plans prepare for a RADV audit?

Plans must prepare for the inevitability of a Risk Adjustment Data Validation (RADV) audit in the future. CMS will send plans a sample of previously submitted encounters and they must provide medical chart documentation to justify the submissions. If they cannot, plans will face major penalties. These penalties would be attributable to all plan revenue based on the percentage of failure in the sample.

Laggard plans need to begin taking seriously the threat that RADV may seriously impact plan revenue and financial stability. A multi-faceted approach will need to be taken, including:

  • Establishing a dedicated RADV team with membership across key departments.
  • Aggressive training of staff in all departments, but principally in finance, medical affairs, compliance, and provider relations, as to the significance of RADV and the audit process.
  • Train providers on the significance of RADV, impact on the plan, and eventually impact to provider reimbursement. The training should include all the published information on invalidity of submissions due to lack of adequate documentation.
  • Conduct mock audits and pro-actively gather medical records from providers each year. This includes an internal process and one with your providers, especially high volume ones.
  • Ensure that claims submission and RADV is part of your delegated oversight process. The plan will be in jeopardy if your vendors and their providers underperform. Smart plans will place delegated subcontractors at some risk for any RADV penalties found on audit.

RADV, EDPS

About The Author

Marc Ryan

Marc S. Ryan serves as MedHOK’s Chief Strategy and Compliance Officer. During his career, Marc has served a number of health plans in executive-level regulatory, compliance, business development, and operations roles. He has launched and operated plans with Medicare, Medicaid, commercial and Exchange lines of business. Marc was the Secretary of Policy and Management and State Budget Director of Connecticut, where he oversaw all aspects of state budgeting and management. In this role, Marc created the state’s Medicaid and SCHIP managed care programs, and oversaw its state employee and retiree health plans. He also created the state’s long-term care continuum program. Marc was nominated by then HHS Secretary Tommy Thompson to serve on a panel of state program experts to advise CMS on aspects of Medicare Part D implementation. He also was nominated by Florida’s Medicaid Secretary to serve on the state’s Medicaid Reform advisory panel. Marc graduated cum laude from the Edmund A. Walsh School of Foreign Service at Georgetown University with a Bachelor of Science in Foreign Service. He received a Master of Public Administration, specializing in local government management and managed healthcare, from the University of New Haven. He was inducted into Sigma Beta Delta, a national honor society for business, management and administration.

0 Comments