The Audit You Won on Evidence but Lost on Formatting
RADV audit submissions follow strict technical specifications. Medical records are submitted through CMS’s Centralized Data Abstraction Tool (CDAT) system in prescribed formats with specific field requirements. Documentation must be associated with the correct enrollee, the correct HCC, and the correct date of service. Files must meet size, format, and naming conventions. Submission must occur within the five-month window through the designated electronic system.
Plans that assemble strong clinical evidence but submit it with technical errors can have records rejected or processed incorrectly. A documentation package with clear MEAT support for a diagnosis that gets rejected because the file format doesn’t meet CDAT specifications is a defensible code that fails on logistics. The clinical evidence existed. The technical submission failed.
This sounds like a minor operational detail. It’s not. When CMS processes hundreds of records per plan across thousands of contracts, technical rejections are processed systematically. A rejected record is treated as missing documentation for the associated HCC. Missing documentation produces a discrepant finding. The plan then has to submit corrected records through a resubmission process that consumes additional time and may not fully recover the lost ground.
Where Technical Failures Concentrate
Three areas produce the most CDAT submission failures. First, file format mismatches. Clinical records come from multiple sources in multiple formats. Some arrive as scanned PDFs, others as EHR exports, others as faxed images. Each must meet CDAT’s specifications. Records that arrive in unsupported formats or exceed size limits need conversion before submission. Plans that don’t build format conversion into their response workflow discover the problem under deadline pressure.
Second, enrollee-HCC association errors. Each submitted record must be correctly linked to the specific enrollee and specific HCC code under review. When plans manage this mapping manually across spreadsheets, transposition errors associate the wrong record with the wrong diagnosis. The clinical evidence may be strong, but it’s attached to the wrong HCC in the submission system.
Third, date-of-service mismatches. CMS allows records from any date within the relevant payment year, but the submitted record must fall within the specified date range. Records from outside the payment year are rejected. Plans that don’t verify date alignment before submission waste evidence packages on records CMS won’t accept.
Building Technical Compliance Into the Workflow
The fix is automating technical validation as a pre-submission checkpoint. Before any record package leaves the plan, the system verifies file format compliance, enrollee-HCC association accuracy, date-of-service alignment, and file size limits. Records that fail any technical check get flagged for correction before submission rather than after rejection.
Plans on unified platforms where coding, evidence preservation, and audit response share the same environment can automate most of this. The system knows which enrollee is associated with which HCC, which records are available for each diagnosis, and which dates fall within the payment year. The technical validation layer applies rules automatically rather than relying on manual checks.
Don’t Lose on Logistics
Plans invest significant resources in building clinical evidence and MEAT-validated coding. Losing that investment to CDAT formatting errors is preventable. Technical compliance in radv audits is a solved problem for plans that build it into their submission workflow. Plans that treat it as an afterthought will keep losing defensible codes to logistics failures that should never have reached the submission stage.
