The Cost Nobody Budgets For
Health plans running risk adjustment on disconnected systems pay an integration tax they rarely quantify. Coding happens in one system. Quality assurance in another. Chart retrieval through a third. RADV audit response in spreadsheets and shared drives. Each system works independently. The cost comes from making them work together. The integration tax shows up in three forms. First, data reconciliation labor. When coding decisions in System A need to be validated against documentation in System B and tracked for audit purposes in System C, someone has to move data between systems, verify it transferred correctly, and reconcile discrepancies. This labor doesn’t produce value. It maintains coherence across an architecture that was never designed to be coherent.
Second, evidence trail gaps. When a coding decision is made in one system and its supporting evidence lives in another, the connection between decision and evidence depends on manual linkage. Manual linkage fails at scale. Some codes lose their evidence connections during transfer. Some evidence packages are incomplete because the QA system doesn’t have access to the AI’s original assessment from the coding system. The gaps are invisible during normal operations and surface during audits, when the plan needs to produce a complete evidence trail and discovers it doesn’t exist in any single system. Third, response time inflation. Every RADV audit response that requires assembling data from multiple systems adds weeks to the five-month timeline. Weeks spent on data assembly are weeks not spent on evidence validation and strategic record selection, the activities that actually determine audit outcomes.
Quantifying the Tax
Plans can estimate their integration tax by measuring three things. How many FTEs spend time moving data between risk adjustment systems? What percentage of audit response time goes to data assembly versus evidence validation? How many evidence trail gaps have been discovered during audit response in the past two years? The FTE count is usually surprising. Data reconciliation work is distributed across multiple roles (analysts, coders, QA reviewers, audit response staff) rather than concentrated in one function, which makes the total labor invisible in standard reporting.
Aggregating the time each role spends on cross-system data management reveals a headcount equivalent that most plans haven’t calculated. Evidence trail gaps are harder to quantify proactively but easy to measure retrospectively. In prior audit responses, how many sampled HCCs required evidence reconstruction because the original evidence chain was broken across systems? Each reconstructed evidence package represents a code that was defensible at submission but became vulnerable during response because the systems didn’t maintain the connection.
What Elimination Looks Like
A unified system where coding, evidence preservation, quality assurance, and audit management share the same data environment eliminates the integration tax by design. There’s no data to reconcile because it was never separated. There are no evidence trail gaps because the trail was built and stored in the same environment. There’s no response time inflation because the audit team queries one system instead of assembling from many.
The Hidden Line Item
Plans comparing the cost of a unified risk adjustment solution against their current multi-system architecture should include the integration tax in the comparison. The license fees for three separate systems may appear lower than one unified platform. Add the labor cost of data reconciliation, the audit exposure from evidence trail gaps, and the response time risk from multi-system assembly, and the total cost of fragmentation routinely exceeds the cost of consolidation. The tax is real. It’s just hidden across enough budget lines that nobody totals it up.