RAPS vs. EDS: Understanding CMS Encounter Data Submission for Risk Adjustment

What Are RAPS and EDS

Medicare Advantage plans must submit clinical data to CMS so the agency can calculate RAF scores and set appropriate capitation payments. For over two decades, two parallel systems have served this purpose: the Risk Adjustment Processing System (RAPS) and the Encounter Data System (EDS). Understanding both systems is essential for any organization involved in risk adjustment.

RAPS was built specifically for risk adjustment. It accepts a limited data set — essentially the minimum information CMS needs to assign HCCs and calculate risk scores. Plans submit beneficiary identifiers, ICD-10 diagnosis codes, provider types, and dates of service. That is the extent of it. No procedure codes, no claim amounts, no detailed service-level data.

EDS operates on an entirely different model. It requires full encounter-level claims data in the 837 standard transaction format — the same format used for commercial and Medicaid claims processing. EDS captures diagnosis codes, procedure codes (CPT and HCPCS), revenue codes, provider NPIs for both rendering and billing providers, place of service, claim charges, and complete member demographic information.

The distinction matters because EDS gives CMS visibility into not just what conditions were documented, but what services were rendered, where they occurred, who performed them, and how much was charged. This level of transparency supports multiple CMS objectives beyond risk adjustment, including payment accuracy audits, quality measurement, and program integrity oversight.

The Transition from RAPS to EDS

CMS began requiring encounter data submission in 2012, running EDS in parallel with RAPS. The agency's long-term goal was always to phase out RAPS entirely, but the transition has been gradual — driven by data quality concerns, plan readiness gaps, and the sheer complexity of changing a data pipeline that processes billions of records annually.

• 2012-2015: CMS mandated EDS submission alongside RAPS but continued to use RAPS as the primary source for risk adjustment calculations. EDS served largely as a data collection mechanism during this period.
• 2016-2019: CMS began incorporating EDS data into risk score calculations using a blended approach. Plans that submitted high-quality EDS data saw diagnoses from both systems counted, while EDS-only diagnoses started contributing to RAF scores.
• 2020-2024: EDS became the dominant source for risk adjustment. CMS applied an EDS-RAPS reconciliation process, using EDS as the primary calculation source and RAPS as a supplemental fallback for specific scenarios.
• 2025-2027: CMS has moved toward full EDS reliance for risk adjustment data processing. RAPS retains limited operational status for legacy corrections and historical data, but new risk adjustment submissions flow through EDS.

This transition timeline matters because plans that invested early in EDS data quality have a structural advantage. Those still treating EDS as secondary to RAPS face increasing risk of data rejection, missed diagnoses, and audit exposure.

Key Differences Between Systems

The gap between RAPS and EDS extends well beyond the volume of data submitted. The systems differ in validation logic, rejection handling, and downstream impact on risk adjustment workflows.

• Data Granularity: RAPS accepts diagnosis-level records. EDS requires full claim-level records with service lines, procedure codes, and financial data. This means EDS encounters must pass the same validation rules as standard claims.
• Validation Strictness: EDS applies front-end edits that RAPS never required. Missing or invalid NPIs, inconsistent dates of service, procedure-to-diagnosis mismatches, and revenue code errors all trigger EDS rejections. Plans frequently see 10-20% higher initial rejection rates when transitioning from RAPS to EDS.
• Correction Mechanisms: RAPS allowed simple chart review submissions and add/delete transactions. EDS requires formal claim replacement or void/resubmission processes that mirror commercial claims correction workflows.
• Audit Traceability: EDS creates a complete audit trail from service to submission. CMS can trace a diagnosis back to the specific encounter, rendering provider, procedure performed, and facility — which is precisely what RADV audits now leverage.
• Submission Deadlines: Both systems operate on similar submission windows, but EDS processing timelines include additional front-end validation steps that effectively shorten the window for successful submission.

EDS Data Requirements

Successful EDS submission requires plans to meet rigorous data quality standards that many organizations initially underestimate. The end-to-end risk adjustment workflow must account for these requirements from the point of care through final submission.

• 837P Professional Claims: Required fields include member ID, rendering provider NPI, billing provider NPI, place of service, CPT/HCPCS procedure codes, ICD-10 diagnosis codes (up to 12 per claim line), dates of service, and charge amounts. Each field must pass CMS front-end validation.
• 837I Institutional Claims: Additional requirements include revenue codes, DRG codes (for inpatient), admission and discharge dates, bill type codes, and attending/operating provider NPIs. Institutional claims have historically higher rejection rates due to the complexity of required fields.
• Provider Data Integrity: Every NPI submitted must be active and valid in the NPPES database. Deceased providers, deactivated NPIs, and provider-type mismatches are common rejection causes that do not surface in RAPS submissions.
• Diagnosis-Procedure Consistency: EDS validates that diagnosis codes are clinically consistent with the procedures billed. A diabetes diagnosis submitted alongside only an orthopedic procedure code may trigger an edit, whereas RAPS would accept the diagnosis without question.
• Timeliness: CMS enforces submission deadlines tied to payment years. Late EDS submissions may not be processed in time for initial risk score calculations, requiring plans to rely on mid-year sweeps or retroactive adjustments.

Impact on Risk Adjustment Revenue

The transition to EDS has measurable revenue implications for Medicare Advantage plans. Organizations that fail to adapt their data submission processes risk losing legitimate RAF score value — not because conditions are absent, but because submission errors prevent CMS from processing them.

• Higher Rejection Rates: Plans transitioning from RAPS-primary to EDS-primary workflows report initial rejection rate increases of 10-25%. Each rejected encounter potentially represents a lost HCC and its associated RAF coefficient.
• Diagnosis Loss from Validation Failures: A diagnosis code that would have been accepted through RAPS may be rejected through EDS due to an unrelated data quality issue — such as an invalid billing NPI or missing revenue code. The diagnosis is clinically valid but technically non-compliant.
• Improved Audit Defensibility: The upside of EDS is that accepted encounters carry stronger audit documentation. Plans that pass EDS validation have already demonstrated claim-level data integrity, which reduces RADV exposure on those encounters.
• Revenue Recovery Opportunities: EDS data enables plans to identify submission gaps more precisely. By analyzing rejection codes and resubmission success rates, risk adjustment analytics teams can prioritize high-value corrections.

Preparing for Full EDS Transition

Plans that have not yet achieved high-quality EDS submission should treat this as a strategic priority. The window for relying on RAPS as a safety net is closing, and CMS has signaled that future risk adjustment enhancements will be built exclusively on EDS infrastructure.

• Conduct an EDS Readiness Assessment: Measure current EDS acceptance rates by claim type, rejection category, and provider network segment. Identify the top five rejection reasons and build targeted remediation plans for each.
• Strengthen Provider Data Management: NPI validation, provider enrollment verification, and credentialing data synchronization are foundational. Many EDS rejections trace back to provider data quality issues that exist upstream of the claims process.
• Implement Pre-Submission Scrubbing: Apply CMS front-end edit logic to encounters before submission. Catching validation failures internally reduces rejection rates and accelerates the diagnosis-to-RAF-score cycle.
• Align Coding and Billing Teams: RAPS-era workflows often separated coding (focused on diagnoses) from billing (focused on claims). EDS requires these functions to work in coordination because the entire claim must be complete and consistent for diagnoses to be accepted.
• Invest in Risk Adjustment Analytics: Real-time visibility into EDS submission status, rejection trends, and resubmission success rates enables proactive management rather than reactive correction.
• Build Resubmission Workflows: Establish clear processes for correcting and resubmitting rejected EDS encounters within CMS deadlines. Timeliness of resubmission directly impacts whether diagnoses count toward the current payment year.