What impact will come with the migration from a RAPS-based RAF score to an encounter data-based RAF score?
The RISE User Group aims to find out the answer to this question. It involves complexity that is difficult to appreciate until you dig in deeper. There are a lot of variables that cause different healthplans to come to different conclusions. Here are some of the key moving parts:
- RAPS uses only 5 data elements to qualify for acceptable data submission while EDPS is aligned with the 5010 claim form of 37 data elements, which presents a lot more hurdles to pass for getting acceptance.
- RAPS excludes a lot of services as ineligible for inclusion (such as labs, skilled nursing, etc.) while EDPS takes it all in. RAPS allows for supplemental data submissions to beef up its comprehensiveness and accuracy. In contrast, EDPS is primarily encounter-based (or claims- based), but EDPS allows retrospective and prospective chart reviews to augment the data from claims.
- Depending on the filtering logic employed by the healthplans for both approaches, there can be a wide divergence in the percentage of diagnoses accepted
In addition to all these variables, we share further challenges getting adequate diagnoses captured. When you take an end-to-end process review, there are several places where robust diagnosis capture gets watered down. “Think of this as a hose”, as Behzad Mohazzebi1 famously said, “and let’s count the leaks”. Some examples:
- The provider office billing systems, complying with the 5010, can only accept 12 ICD codes
- The 837 submission abides by this restriction and, in order to gain a full set of diagnoses for truly complex patients, multiple 837s would be required. However, that sets up claims processing duplicate flags at the payers’ end and would also require some special programming when submitting more than 12 ICD codes to EDPS as plans would be required to LINK claims together.
- The handoff to billing services often result in truncating diagnoses, as do clearinghouses and other intermediate hands in the process
- Hospital billing systems can often truncate the rich set of diagnoses that their coders extract from medical records
- The claims systems at healthplans and insurance companies also have been found to frequently drop diagnoses because they are old systems that were not designed to accept a large number of diagnostic codes
The User Group hopes to begin tracking and reporting as testing of EDS submissions progresses. We are expecting to gauge the level of data rejection, the types of filtering logic pursued by the plans, and root causes discovered in the process. We are publishing a worksheet that is being used as the initial data collection tool for any RISE members to view. Whatever we find as we work on this, we plan to publish and share in a transparent manner. All Medicare Advantage Organizations are invited to join the group. However, it may also be of interest to those with ACA-Marketplace products, since that will be entirely an encounter-based risk adjustment program.
Download the Excel Spreadsheet we are using for data collection: click here to download first page second page third page
1 Behzad Mohazzebi, founder of DCA, now part of Altegra Health