Coder hours
Spent on chart orientation and MEAT documentation lookups instead of coding judgment.
HCC Risk Adjustment. The closed loop risk adjustment was missing.
Prospective capture and retrospective recovery on one Operating System. Chart orientation and MEAT documentation lookups come off your coders' desks. Every code carries its evidence chain. RADV ready by default.
The problem
The playbook you've been handed.
Risk-bearing organizations have been trained to run risk adjustment as two disconnected programs. Prospective lives with the clinical team. Retrospective lives with the coding vendor. The chart chase is its own line item. RADV gets called in when the audit notice arrives.
By the time it's all stood up, you don't have a risk adjustment strategy. You have four programs that don't share a chart, three vendors that don't share an audit trail, and a coder workforce that spends more time orienting to cases than coding them.
Here is where it breaks: the model assumes prospective and retrospective are different problems. They are the same problem on different time axes. Every HCC missed at the visit becomes a chase letter, a chart pull, and a retrospective code. Every retrospective finding is a signal that prospective workflows are not catching what the chart already supports.
01
Premium calendar starts.
Prospective program launches with one tool, one vendor, one workflow.
02
Retrospective sweep begins.
A second tool, a second vendor, a second workflow. Different evidence model. Different audit trail.
03
RADV notice arrives.
A third workflow. Charts pulled, evidence reassembled, codes defended one at a time.
04
Repeat next year.
Every payment year. Every audit cycle. Every CMS model update.
It is not the coder budget. It is what happens between the seams.
When prospective, retrospective, CDI, and RADV all run as separate programs on separate platforms, the same chart gets pulled three times and the audit trail lives in three different vendor portals.
The lightbulb
Prospective and retrospective are the same workflow on different time axes.
Prospective vendors optimize for the visit. Retrospective vendors optimize for the sweep. Nothing in the current stack closes the loop between what gets missed prospectively and what gets recovered retrospectively. The answer is a single Operating System where prospective capture and retrospective recovery operate as one closed loop.
Engine 01
Prospective capture.
Every supportable HCC surfaced at the visit with chart evidence ready for the clinician.
Closed loop
Engine 02
Retrospective recovery.
Every chart reviewed, every missed condition recovered, every code defended.
Each retrospective finding feeds back to prospective workflows so the same condition does not slip through next year.
Prospective risk adjustment · Capture at the visit
Catch supportable HCCs before the encounter closes.
Candidate conditions surfaced with chart evidence and supporting ICD-10 codes. Documentation gaps route a CDI query before the visit is signed.
- 01
Chart reviewed.
Pre-visit and in-visit clinical notes read continuously.
- 02
HCCs surfaced.
Supportable conditions identified with chart evidence and ICD-10 codes attached.
- 03
CDI query routed.
Documentation gaps generate queries to the responsible clinician. Closure tracked.
- 04
Coder confirms.
Coder reviews captured codes with evidence in context before submission.
Retrospective risk adjustment · Recovery and defense
Every chart reviewed. RADV ready by default.
Every chart reviewed for supportable HCCs that were not captured prospectively. MEAT evidence assembled with every recommendation. RADV readiness is the default state because every code carries its evidence chain.
- 01
Charts mined.
Every chart in the population reviewed. Every supportable code surfaced.
- 02
Evidence assembled.
MEAT documentation organized. Chart excerpts attached. Reasoning documented.
- 03
Coder confirms.
Coder reviews each candidate with evidence in context. Confirms or rejects.
- 04
RADV ready.
Every confirmed code carries its evidence chain. Auditors see what they need in one view.
Full risk adjustment lifecycle. Defensible at every step.
Prospective and retrospective on one Operating System.
Same chart. Same evidence model. Same audit trail. Findings from one workflow teach the other.
Specificity-first capture.
Ai Workers surface what the chart supports, not what the model guesses. Coder review is mandatory on every code.
Your coders make the call.
Ai Workers handle extraction and MEAT organization. Your coders and CDI specialists own every code.
RADV-ready provenance.
Every code carries the chart excerpt, reasoning path, and coder sign-off in one view. RADV prep stops being a quarterly fire drill.
The Glassbox. Every code, explained.
Every captured code arrives with the supporting documentation, MEAT evidence, ICD-10 mapping, and reasoning path visible. No black box. The art of medicine and the discipline of coding both stay with your team.
Full audit trail
Every code, query, and coder action documented with clinical reasoning and evidence references. RADV ready at all times.
CMS-HCC and HHS-HCC coverage
Medicare Advantage, ACA marketplace, and commercial risk adjustment models supported with appropriate policy configuration.
Configurable coding guidelines
Plan-specific guidelines, internal standards, and CMS model updates applied consistently across both workflows.
Human in the loop
Ai Workers prepare the work. Your coders and CDI specialists apply their expertise and make the call.
Continuous learning
Every confirmation, override, and audit outcome strengthens future recommendations across both workflows.
The Glassbox
Every captured code arrives with the supporting documentation, MEAT evidence, ICD-10 mapping, and reasoning path visible. No black box.
The next time someone tells you the answer to your capture gap is another retrospective sweep or another prospective tool, ask what that delivers for your audit posture this year.
Risk adjustment should give your coders their time back. It should give your CDI team queries that close. It should give your compliance leaders an audit defense that already exists, not one that has to be assembled when the notice arrives.