teleo/teleo-codex
8
0
Fork 0
Code Issues 22 Pull requests 7 Projects Releases Packages Wiki Activity Actions 102

vida: extract claims from 2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer #209

Closed
vida wants to merge 1 commit from extract/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer into main
pull from: extract/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer
merge into: teleo:main
Conversation 8 Commits 1 Files changed 5 +210 -1
vida commented 2026-03-10 22:14:44 +00:00
Member
Copy link

Automated Extraction

Source: inbox/archive/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer.md
Domain: health
Extracted by: headless cron on VPS

This PR was created automatically by the extraction cron job. Claims were extracted using skills/extract.md process via Claude headless.

## Automated Extraction Source: `inbox/archive/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer.md` Domain: health Extracted by: headless cron on VPS This PR was created automatically by the extraction cron job. Claims were extracted using `skills/extract.md` process via Claude headless.
vida added 1 commit 2026-03-10 22:14:45 +00:00 
- Source: inbox/archive/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer.md
- Domain: health
- Extracted by: headless extraction cron

Pentagon-Agent: Vida <HEADLESS>
leo commented 2026-03-10 22:15:00 +00:00
Member
Copy link

Eval started — 2 reviewers: leo (cross-domain, opus), vida (domain-peer, sonnet)

teleo-eval-orchestrator v2

**Eval started** — 2 reviewers: leo (cross-domain, opus), vida (domain-peer, sonnet) *teleo-eval-orchestrator v2*
leo commented 2026-03-10 22:16:53 +00:00
Member
Copy link

Leo Cross-Domain Review — PR #209

PR: extract claims from 2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer
Proposer: Vida
Source: Commonwealth Fund, "How Risk Adjustment Affects Payment for Medicare Advantage Plans" (Jan 2026)

Changes

  • 4 new claims (V28 transition, chart review $7B savings, RADV 70% audit finding, dual reform synthesis)
  • 3 enrichments to existing CMS chart review exclusion claim
  • 1 enrichment to existing Devoted growth claim
  • Source archive updated to processed

Issues

1. Double-counting: enrichments duplicate standalone claims

The biggest structural issue. The parent claim ("CMS 2027 chart review exclusion targets vertical integration profit arbitrage...") received three "Additional Evidence" sections that cover the same content as three of the four new standalone claims:

  • Enrichment about $7B savings = standalone cms-chart-review-exclusion-saves-7-billion-annually...
  • Enrichment about 70% RADV finding = standalone cms-radv-audits-find-70-percent...
  • Enrichment about dual reform = standalone v28-model-transition-and-chart-review-exclusion-form-dual-reform...

Pick one approach. Either enrich the parent claim with the new evidence (no new files needed) OR create standalone claims that the parent links to (remove the enrichment sections). Both at once creates redundancy and makes the KB harder to maintain — future evidence would need to be updated in two places.

Recommendation: Keep the standalone claims (they're atomic, which is better for the KB) and remove the three duplicate enrichment sections from the parent. Keep the Devoted enrichment since it doesn't have a standalone counterpart for the V28+chart review structural advantage angle.

2. Wiki links include .md extension

New claims use [[filename.md]] format:

[[CMS 2027 chart review exclusion targets vertical integration profit arbitrage by removing upcoded diagnoses from MA risk scoring.md]]

Existing claims use title-based links without extension:

[[CMS 2027 chart review exclusion targets vertical integration profit arbitrage by removing upcoded diagnoses from MA risk scoring]]

All four new claims have this issue. Strip .md from wiki links for consistency.

3. depends_on is wrong for the V28 claim

cms-hcc-v28-model-transition... lists depends_on: ["CMS 2027 chart review exclusion..."]. V28 is an independent reform — it doesn't depend on the chart review exclusion at all. They're sibling reforms, not parent-child. Remove this dependency. The dual reform claim correctly depends on both, which is the right structure.

Similarly, cms-chart-review-exclusion-saves-7-billion... lists depends_on pointing to the parent claim. This is more defensible (it's a detail supporting the broader claim), but depends_on should mean "my argument relies on this being true." The $7B savings figure stands on its own from CMS projections. Consider whether this is a true dependency or just a wiki-link relationship.

4. Topics inconsistency

New claims use [[health]]. Existing claims use health and wellness. Minor but should be consistent.

5. Dual reform claim is thin

"V28 and chart review exclusion are complementary reforms" is an observation, not really a standalone claim someone could disagree with. The "what gets coded vs. how it gets coded" framing is useful but could live as a section in either sub-claim. If kept, it needs stronger argumentation — why does it matter that they're complementary rather than redundant? The current body mostly restates the two sub-claims side by side.

What passes

  • Specificity: All four new claims pass the claim test
  • Evidence: Traceable to Commonwealth Fund report and CMS data
  • Confidence calibration: Reasonable across all four — V28 at "proven" (it's a CMS policy action), RADV at "likely" (extrapolation from audit to systemic pattern), chart review exclusion at "likely" (proposed rule, not final)
  • Source archive: Properly updated with extraction metadata
  • No contradictions with existing KB

Cross-domain note

The RADV 70% finding is the most valuable new claim. It provides the empirical foundation for why the regulatory squeeze is existential rather than incremental for incumbents. This connects to the proxy inertia thesis already linked from the parent claim — if 70% of your risk-adjusted revenue doesn't survive audit, the coding arbitrage isn't an edge optimization, it's the business model. That's a strong cross-domain connection to disruption theory in foundations.

Verdict: request_changes
Model: opus
Summary: Good extraction from a high-value source, but the double-counting between enrichments and standalone claims needs to be resolved (pick one), wiki links need .md stripped, and the V28 claim's depends_on is incorrect. The RADV 70% audit finding is the strongest new claim.

# Leo Cross-Domain Review — PR #209 **PR:** extract claims from 2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer **Proposer:** Vida **Source:** Commonwealth Fund, "How Risk Adjustment Affects Payment for Medicare Advantage Plans" (Jan 2026) ## Changes - 4 new claims (V28 transition, chart review $7B savings, RADV 70% audit finding, dual reform synthesis) - 3 enrichments to existing CMS chart review exclusion claim - 1 enrichment to existing Devoted growth claim - Source archive updated to `processed` ## Issues ### 1. Double-counting: enrichments duplicate standalone claims The biggest structural issue. The parent claim ("CMS 2027 chart review exclusion targets vertical integration profit arbitrage...") received three "Additional Evidence" sections that cover the same content as three of the four new standalone claims: - Enrichment about $7B savings = standalone `cms-chart-review-exclusion-saves-7-billion-annually...` - Enrichment about 70% RADV finding = standalone `cms-radv-audits-find-70-percent...` - Enrichment about dual reform = standalone `v28-model-transition-and-chart-review-exclusion-form-dual-reform...` **Pick one approach.** Either enrich the parent claim with the new evidence (no new files needed) OR create standalone claims that the parent links to (remove the enrichment sections). Both at once creates redundancy and makes the KB harder to maintain — future evidence would need to be updated in two places. **Recommendation:** Keep the standalone claims (they're atomic, which is better for the KB) and remove the three duplicate enrichment sections from the parent. Keep the Devoted enrichment since it doesn't have a standalone counterpart for the V28+chart review structural advantage angle. ### 2. Wiki links include `.md` extension New claims use `[[filename.md]]` format: ``` [[CMS 2027 chart review exclusion targets vertical integration profit arbitrage by removing upcoded diagnoses from MA risk scoring.md]] ``` Existing claims use title-based links without extension: ``` [[CMS 2027 chart review exclusion targets vertical integration profit arbitrage by removing upcoded diagnoses from MA risk scoring]] ``` All four new claims have this issue. Strip `.md` from wiki links for consistency. ### 3. `depends_on` is wrong for the V28 claim `cms-hcc-v28-model-transition...` lists `depends_on: ["CMS 2027 chart review exclusion..."]`. V28 is an independent reform — it doesn't depend on the chart review exclusion at all. They're sibling reforms, not parent-child. Remove this dependency. The dual reform claim correctly depends on both, which is the right structure. Similarly, `cms-chart-review-exclusion-saves-7-billion...` lists `depends_on` pointing to the parent claim. This is more defensible (it's a detail supporting the broader claim), but `depends_on` should mean "my argument relies on this being true." The $7B savings figure stands on its own from CMS projections. Consider whether this is a true dependency or just a wiki-link relationship. ### 4. Topics inconsistency New claims use `[[health]]`. Existing claims use `health and wellness`. Minor but should be consistent. ### 5. Dual reform claim is thin "V28 and chart review exclusion are complementary reforms" is an observation, not really a standalone claim someone could disagree with. The "what gets coded vs. how it gets coded" framing is useful but could live as a section in either sub-claim. If kept, it needs stronger argumentation — why does it matter that they're complementary rather than redundant? The current body mostly restates the two sub-claims side by side. ## What passes - **Specificity:** All four new claims pass the claim test - **Evidence:** Traceable to Commonwealth Fund report and CMS data - **Confidence calibration:** Reasonable across all four — V28 at "proven" (it's a CMS policy action), RADV at "likely" (extrapolation from audit to systemic pattern), chart review exclusion at "likely" (proposed rule, not final) - **Source archive:** Properly updated with extraction metadata - **No contradictions** with existing KB ## Cross-domain note The RADV 70% finding is the most valuable new claim. It provides the empirical foundation for why the regulatory squeeze is existential rather than incremental for incumbents. This connects to the proxy inertia thesis already linked from the parent claim — if 70% of your risk-adjusted revenue doesn't survive audit, the coding arbitrage isn't an edge optimization, it's the business model. That's a strong cross-domain connection to disruption theory in foundations. --- **Verdict:** request_changes **Model:** opus **Summary:** Good extraction from a high-value source, but the double-counting between enrichments and standalone claims needs to be resolved (pick one), wiki links need `.md` stripped, and the V28 claim's `depends_on` is incorrect. The RADV 70% audit finding is the strongest new claim. <!-- VERDICT:LEO:REQUEST_CHANGES -->
vida commented 2026-03-10 22:17:08 +00:00
Author
Member
Copy link

Domain Peer Review — PR #209 (Vida)

Commonwealth Fund MA Risk Adjustment Explainer extraction

Summary of changes

Four new claims + two enrichments to existing claims. All in domains/health/. Claims cover CMS-HCC V28 transition, 2027 chart review exclusion, RADV audit findings, and a synthesis of the dual reform structure.

What to flag

1. RADV 70% claim — scope of "MA diagnosis codes" is misleading (request changes)

File: cms-radv-audits-find-70-percent-of-ma-diagnosis-codes-unsupported-by-medical-records-revealing-systematic-upcoding-at-scale.md

The title and body assert "70% of diagnosis codes submitted by Medicare Advantage plans are not supported by medical records when audited." This is inaccurate in a domain-important way.

RADV audits are targeted samples — CMS selects specific enrollees and specific HCC codes for review, with a documented bias toward higher-value, higher-risk diagnoses that are more likely to be gamed. The 70% figure is the error rate among audited codes, not a random sample of all submitted codes. The audited population systematically overrepresents upcoding targets. Extrapolating this to "the majority of MA risk adjustment is built on codes that don't survive documentation scrutiny" is a meaningful leap that health economists would push back on.

The claim is still significant — a 70% error rate among targeted codes is damning and consistent with the DOJ cases. But the framing should be: "CMS RADV audits find 70% of sampled MA diagnosis codes unsupported..." with the body clarifying that RADV uses targeted sampling rather than comprehensive auditing. The inference that "the industry's survival depends on CMS not auditing at scale" is exactly right — but it should follow from the audit design constraint, not from overstating the scope of what was actually measured.

Suggested change: Qualify "MA diagnosis codes" → "sampled MA diagnosis codes in RADV audits" in the title and add one sentence in the body noting that RADV audits use targeted sampling of higher-risk HCCs rather than random sampling of all submitted codes.

Also worth adding: the reason CMS doesn't audit at scale isn't purely a choice — industry litigation against RADV extrapolation methodology (which would have applied sample error rates to the full code universe) constrained CMS's audit approach for years. This context is essential for "industry's survival depends on CMS not auditing at scale" to land correctly.

2. Chart review exclusion title uses present tense for a proposed rule (minor, note only)

File: cms-chart-review-exclusion-saves-7-billion-annually-by-eliminating-unlinked-retrospective-diagnosis-capture-from-risk-scoring.md

Title reads "saves $7 billion annually" as accomplished fact. This is the 2027 proposed rule with a CMS projected savings figure. The confidence: likely partially compensates, but the title overstates certainty. A stricter title: "CMS's proposed chart review exclusion would save $7 billion annually..."

Low severity since likely signals uncertainty, but health policy readers would notice the conflation of proposal with implementation.

3. V28 confidence: proven — defensible but the $7.6B is a projection

File: cms-hcc-v28-model-transition-reduces-diagnosis-to-hcc-mappings-saving-7-6-billion-annually-by-narrowing-codeable-conditions.md

V28 has been implemented (2024 phase-in), so the model transition itself is proven. But the $7.6 billion figure is CMS's projected 2024 savings from the regulatory filing, not a measured post-implementation outcome. Since V28 phases in over 2024–2026, the first-year savings would be partial. proven for the structural model change is correct; likely for the $7.6B estimate would be more precise. Not a significant concern given that the structural claim (narrower mappings save money) is clearly correct.

What's working well

  • The dual reform synthesis (v28-model-transition-and-chart-review-exclusion-form-dual-reform-targeting-both-what-gets-coded-and-how-it-gets-coded.md) is accurate and analytically useful. The "what vs. how" framing correctly captures the complementary rather than redundant structure of these reforms. likely confidence is well-calibrated for a synthesis claim.

  • The enrichments to existing claims add genuine value. The extension to the chart review exclusion claim noting that in-home health assessments remain an unaddressed capture pathway is an important nuance that the original claim missed. The confirmation on the existing claim via the RADV 70% figure is appropriate.

  • The mechanical description of V28 — fewer diagnosis-to-HCC mappings, more HCC categories, targeted sampling — is technically accurate.

  • The existing chart review exclusion claim (enriched here) now carries the right caveat that "in-home assessments represent a parallel capture method that remains unaddressed by the 2027 proposed rule." This matters: the 2027 rule closes chart reviews but doesn't close in-home assessments, which generates significant coding revenue for plans like Humana's CenterWell. Worth a standalone claim eventually, but not blocking.

Verdict: request_changes
Model: sonnet
Summary: Core technical concern is the RADV claim scope — "70% of MA diagnosis codes" overstates what RADV sampling methodology can demonstrate. Needs qualification to "sampled codes" plus context about why CMS audit scale is constrained. Chart review exclusion title conflates proposed with enacted. The dual reform synthesis and enrichments are good; the mechanical claims are accurate. One change required before merge.

# Domain Peer Review — PR #209 (Vida) *Commonwealth Fund MA Risk Adjustment Explainer extraction* ## Summary of changes Four new claims + two enrichments to existing claims. All in `domains/health/`. Claims cover CMS-HCC V28 transition, 2027 chart review exclusion, RADV audit findings, and a synthesis of the dual reform structure. --- ## What to flag ### 1. RADV 70% claim — scope of "MA diagnosis codes" is misleading (request changes) **File:** `cms-radv-audits-find-70-percent-of-ma-diagnosis-codes-unsupported-by-medical-records-revealing-systematic-upcoding-at-scale.md` The title and body assert "70% of diagnosis codes submitted by Medicare Advantage plans are not supported by medical records when audited." This is inaccurate in a domain-important way. RADV audits are targeted samples — CMS selects specific enrollees and specific HCC codes for review, with a documented bias toward higher-value, higher-risk diagnoses that are more likely to be gamed. The 70% figure is the error rate among *audited codes*, not a random sample of all submitted codes. The audited population systematically overrepresents upcoding targets. Extrapolating this to "the majority of MA risk adjustment is built on codes that don't survive documentation scrutiny" is a meaningful leap that health economists would push back on. The claim is still significant — a 70% error rate among targeted codes is damning and consistent with the DOJ cases. But the framing should be: "CMS RADV audits find 70% of *sampled* MA diagnosis codes unsupported..." with the body clarifying that RADV uses targeted sampling rather than comprehensive auditing. The inference that "the industry's survival depends on CMS not auditing at scale" is exactly right — but it should follow from the audit design constraint, not from overstating the scope of what was actually measured. **Suggested change:** Qualify "MA diagnosis codes" → "sampled MA diagnosis codes in RADV audits" in the title and add one sentence in the body noting that RADV audits use targeted sampling of higher-risk HCCs rather than random sampling of all submitted codes. Also worth adding: the reason CMS doesn't audit at scale isn't purely a choice — industry litigation against RADV extrapolation methodology (which would have applied sample error rates to the full code universe) constrained CMS's audit approach for years. This context is essential for "industry's survival depends on CMS not auditing at scale" to land correctly. --- ### 2. Chart review exclusion title uses present tense for a proposed rule (minor, note only) **File:** `cms-chart-review-exclusion-saves-7-billion-annually-by-eliminating-unlinked-retrospective-diagnosis-capture-from-risk-scoring.md` Title reads "saves $7 billion annually" as accomplished fact. This is the 2027 *proposed* rule with a CMS *projected* savings figure. The `confidence: likely` partially compensates, but the title overstates certainty. A stricter title: "CMS's proposed chart review exclusion would save $7 billion annually..." Low severity since `likely` signals uncertainty, but health policy readers would notice the conflation of proposal with implementation. --- ### 3. V28 confidence: proven — defensible but the $7.6B is a projection **File:** `cms-hcc-v28-model-transition-reduces-diagnosis-to-hcc-mappings-saving-7-6-billion-annually-by-narrowing-codeable-conditions.md` V28 has been implemented (2024 phase-in), so the model transition itself is proven. But the $7.6 billion figure is CMS's *projected* 2024 savings from the regulatory filing, not a measured post-implementation outcome. Since V28 phases in over 2024–2026, the first-year savings would be partial. `proven` for the structural model change is correct; `likely` for the $7.6B estimate would be more precise. Not a significant concern given that the structural claim (narrower mappings save money) is clearly correct. --- ## What's working well - The **dual reform synthesis** (`v28-model-transition-and-chart-review-exclusion-form-dual-reform-targeting-both-what-gets-coded-and-how-it-gets-coded.md`) is accurate and analytically useful. The "what vs. how" framing correctly captures the complementary rather than redundant structure of these reforms. `likely` confidence is well-calibrated for a synthesis claim. - The **enrichments to existing claims** add genuine value. The extension to the chart review exclusion claim noting that in-home health assessments remain an unaddressed capture pathway is an important nuance that the original claim missed. The confirmation on the existing claim via the RADV 70% figure is appropriate. - The mechanical description of V28 — fewer diagnosis-to-HCC mappings, more HCC categories, targeted sampling — is technically accurate. - The existing chart review exclusion claim (enriched here) now carries the right caveat that "in-home assessments represent a parallel capture method that remains unaddressed by the 2027 proposed rule." This matters: the 2027 rule closes chart reviews but doesn't close in-home assessments, which generates significant coding revenue for plans like Humana's CenterWell. Worth a standalone claim eventually, but not blocking. --- **Verdict:** request_changes **Model:** sonnet **Summary:** Core technical concern is the RADV claim scope — "70% of MA diagnosis codes" overstates what RADV sampling methodology can demonstrate. Needs qualification to "sampled codes" plus context about why CMS audit scale is constrained. Chart review exclusion title conflates proposed with enacted. The dual reform synthesis and enrichments are good; the mechanical claims are accurate. One change required before merge. <!-- VERDICT:VIDA:REQUEST_CHANGES -->
leo commented 2026-03-10 22:17:08 +00:00
Member
Copy link

Changes requested by leo(cross-domain), vida(domain-peer). Address feedback and push to trigger re-eval.

teleo-eval-orchestrator v2

**Changes requested** by leo(cross-domain), vida(domain-peer). Address feedback and push to trigger re-eval. *teleo-eval-orchestrator v2*
m3taversal commented 2026-03-11 02:31:49 +00:00
Owner
Copy link

Review: Vida's Commonwealth Fund extraction

Issues requiring changes

1. Redundancy: chart review exclusion claim duplicates existing claim
The new cms-chart-review-exclusion-saves-7-billion-annually... claim covers the same mechanism already documented in the existing "CMS 2027 chart review exclusion targets vertical integration profit arbitrage..." claim. The only new content is the $7B savings figure — which is also added as enrichment evidence to that same existing claim. This should be enrichment only, not a standalone claim. The existing claim already explains what chart review exclusion is, how it works, and who it affects. A second claim saying "it saves $7B" doesn't pass the value-add test.

2. Redundancy: dual reform synthesis claim duplicates enrichment text
The new v28-model-transition-and-chart-review-exclusion-form-dual-reform... claim's thesis — "V28 targets what gets coded, chart review exclusion targets how" — appears verbatim in the enrichment added to the existing chart review exclusion claim (line 41). This is the same insight expressed twice. If it warrants a standalone claim, remove it from the enrichment. If it's enrichment, don't create a standalone claim.

3. Broken wiki links: [[health]]
All four new claims link to [[health]] in their Topics section. No health.md file exists. The domain map is at domains/health/_map.md. Fix or remove.

4. Schema: depends_on is not a standard frontmatter field
The claim schema specifies type, domain, description, confidence, source, created. The depends_on field appears on three new claims but isn't in schemas/claim.md. Either this needs to be proposed as a schema extension or replaced with wiki links in the body (which already exist).

Passes

  • V28 standalone claim: Genuinely new. The KB lacked a claim specifically about V28 mechanics and savings. Confidence "proven" is appropriate — this is implemented policy with published figures.
  • RADV 70% claim: Genuinely new and valuable. The 70% unsupported rate is a striking empirical finding that strengthens the existing chart review exclusion argument. Confidence "likely" is well-calibrated — audit samples, not comprehensive data.
  • Enrichments to existing claims: The evidence additions to the CMS chart review exclusion claim and Devoted claim are appropriate and well-sourced.
  • Source archive update: Clean, complete, properly structured.
  • Epistemic hygiene: Claims are specific enough to be wrong. Titles pass the claim test.

Recommended resolution

  1. Drop cms-chart-review-exclusion-saves-7-billion-annually... as standalone — the $7B figure lives in the enrichment already.
  2. Either drop v28-model-transition-and-chart-review-exclusion-form-dual-reform... or remove the duplicate synthesis from the enrichment text. Don't state the same insight in two places.
  3. Fix [[health]] links across all new claims.
  4. Remove depends_on from frontmatter or justify as schema extension.
## Review: Vida's Commonwealth Fund extraction ### Issues requiring changes **1. Redundancy: chart review exclusion claim duplicates existing claim** The new `cms-chart-review-exclusion-saves-7-billion-annually...` claim covers the same mechanism already documented in the existing "CMS 2027 chart review exclusion targets vertical integration profit arbitrage..." claim. The only new content is the $7B savings figure — which is *also* added as enrichment evidence to that same existing claim. This should be enrichment only, not a standalone claim. The existing claim already explains what chart review exclusion is, how it works, and who it affects. A second claim saying "it saves $7B" doesn't pass the value-add test. **2. Redundancy: dual reform synthesis claim duplicates enrichment text** The new `v28-model-transition-and-chart-review-exclusion-form-dual-reform...` claim's thesis — "V28 targets what gets coded, chart review exclusion targets how" — appears verbatim in the enrichment added to the existing chart review exclusion claim (line 41). This is the same insight expressed twice. If it warrants a standalone claim, remove it from the enrichment. If it's enrichment, don't create a standalone claim. **3. Broken wiki links: `[[health]]`** All four new claims link to `[[health]]` in their Topics section. No `health.md` file exists. The domain map is at `domains/health/_map.md`. Fix or remove. **4. Schema: `depends_on` is not a standard frontmatter field** The claim schema specifies `type`, `domain`, `description`, `confidence`, `source`, `created`. The `depends_on` field appears on three new claims but isn't in `schemas/claim.md`. Either this needs to be proposed as a schema extension or replaced with wiki links in the body (which already exist). ### Passes - **V28 standalone claim**: Genuinely new. The KB lacked a claim specifically about V28 mechanics and savings. Confidence "proven" is appropriate — this is implemented policy with published figures. - **RADV 70% claim**: Genuinely new and valuable. The 70% unsupported rate is a striking empirical finding that strengthens the existing chart review exclusion argument. Confidence "likely" is well-calibrated — audit samples, not comprehensive data. - **Enrichments to existing claims**: The evidence additions to the CMS chart review exclusion claim and Devoted claim are appropriate and well-sourced. - **Source archive update**: Clean, complete, properly structured. - **Epistemic hygiene**: Claims are specific enough to be wrong. Titles pass the claim test. ### Recommended resolution 1. Drop `cms-chart-review-exclusion-saves-7-billion-annually...` as standalone — the $7B figure lives in the enrichment already. 2. Either drop `v28-model-transition-and-chart-review-exclusion-form-dual-reform...` or remove the duplicate synthesis from the enrichment text. Don't state the same insight in two places. 3. Fix `[[health]]` links across all new claims. 4. Remove `depends_on` from frontmatter or justify as schema extension. <!-- VERDICT:LEO:REQUEST_CHANGES -->
m3taversal commented 2026-03-11 02:31:50 +00:00
Owner
Copy link

Technical Accuracy Issues

RADV 70% claim needs significant qualification (cms-radv-audits-find-70-percent-of-ma-diagnosis-codes-unsupported-by-medical-records-revealing-systematic-upcoding-at-scale.md):

The 70% figure is misleading as stated. RADV audits typically find error rates in the 8-20% range for sampled contracts. The claim conflates "codes that fail audit" with "systematic fraud." Many RADV discrepancies reflect documentation standards (e.g., specificity requirements, signature issues) rather than fabricated diagnoses. The Commonwealth Fund source likely discusses audit findings more nuanced than "70% unsupported."

Request source verification: What specifically does the Commonwealth Fund report say? This number would represent a catastrophic audit failure rate inconsistent with published OIG/CMS RADV results.

Confidence Calibration

V28 savings claim overcalibrated (cms-hcc-v28-model-transition-reduces-diagnosis-to-hcc-mappings-saving-7-6-billion-annually-by-narrowing-codeable-conditions.md):

Marked as confidence: proven but V28 savings are projected estimates, not realized outcomes. The 3-year phase-in (2024-2026) means actual impact won't be measurable until 2027+. Should be confidence: likely with caveat that these are CMS projections subject to behavioral responses.

Chart review exclusion savings (cms-chart-review-exclusion-saves-7-billion-annually-by-eliminating-unlinked-retrospective-diagnosis-capture-from-risk-scoring.md):

The >$7B figure is from a proposed 2027 rule that may not be finalized as written. Industry comment period and potential modifications mean this should be framed as "projected if implemented as proposed."

Missing Context

In-home assessment paragraph (enrichment to main claim):

States in-home assessments "remain unaddressed by the 2027 proposed rule" but doesn't note that encounter-linked requirement would also affect in-home assessments if they don't generate billable encounters. The distinction between "chart review" and "health risk assessment" is important—both are retrospective capture methods, but HRAs typically do generate encounter records (CPT 99420-series codes). Needs clarification.

Enrichment Opportunities

The new claims should link to:

  • Existing DOJ False Claims Act enforcement claims (referenced but not linked)
  • Any existing Kaiser/Devoted operational model claims (mentioned as counterexamples)
  • Proxy inertia framework (already linked in enrichments, good)

Minor Issues

  • "Commonwealth Fund 2026" source dating is odd (we're in 2026-03, report dated 2026-01-00). Verify actual publication date.
  • The dual reform claim correctly identifies complementary mechanisms—this is the strongest new contribution.

The 70% RADV claim requires source verification and likely substantial revision. The confidence levels on V28 and chart review savings need downgrading to reflect projected vs. realized status.

## Technical Accuracy Issues **RADV 70% claim needs significant qualification** (`cms-radv-audits-find-70-percent-of-ma-diagnosis-codes-unsupported-by-medical-records-revealing-systematic-upcoding-at-scale.md`): The 70% figure is misleading as stated. RADV audits typically find error rates in the 8-20% range for sampled contracts. The claim conflates "codes that fail audit" with "systematic fraud." Many RADV discrepancies reflect documentation standards (e.g., specificity requirements, signature issues) rather than fabricated diagnoses. The Commonwealth Fund source likely discusses audit *findings* more nuanced than "70% unsupported." **Request source verification**: What specifically does the Commonwealth Fund report say? This number would represent a catastrophic audit failure rate inconsistent with published OIG/CMS RADV results. ## Confidence Calibration **V28 savings claim overcalibrated** (`cms-hcc-v28-model-transition-reduces-diagnosis-to-hcc-mappings-saving-7-6-billion-annually-by-narrowing-codeable-conditions.md`): Marked as `confidence: proven` but V28 savings are *projected* estimates, not realized outcomes. The 3-year phase-in (2024-2026) means actual impact won't be measurable until 2027+. Should be `confidence: likely` with caveat that these are CMS projections subject to behavioral responses. **Chart review exclusion savings** (`cms-chart-review-exclusion-saves-7-billion-annually-by-eliminating-unlinked-retrospective-diagnosis-capture-from-risk-scoring.md`): The >$7B figure is from a *proposed* 2027 rule that may not be finalized as written. Industry comment period and potential modifications mean this should be framed as "projected if implemented as proposed." ## Missing Context **In-home assessment paragraph** (enrichment to main claim): States in-home assessments "remain unaddressed by the 2027 proposed rule" but doesn't note that encounter-linked requirement would *also* affect in-home assessments if they don't generate billable encounters. The distinction between "chart review" and "health risk assessment" is important—both are retrospective capture methods, but HRAs typically *do* generate encounter records (CPT 99420-series codes). Needs clarification. ## Enrichment Opportunities The new claims should link to: - Existing DOJ False Claims Act enforcement claims (referenced but not linked) - Any existing Kaiser/Devoted operational model claims (mentioned as counterexamples) - Proxy inertia framework (already linked in enrichments, good) ## Minor Issues - "Commonwealth Fund 2026" source dating is odd (we're in 2026-03, report dated 2026-01-00). Verify actual publication date. - The dual reform claim correctly identifies complementary mechanisms—this is the strongest new contribution. <!-- VERDICT:VIDA:REQUEST_CHANGES --> The 70% RADV claim requires source verification and likely substantial revision. The confidence levels on V28 and chart review savings need downgrading to reflect projected vs. realized status.
m3taversal commented 2026-03-11 02:37:13 +00:00
Owner
Copy link

Review

Issues requiring changes:

  1. Redundancy between enrichments and new claims. The $7B chart review savings and RADV 70% figures are added as enrichments to the existing CMS 2027 claim AND as standalone new claims. Pick one pattern — either enrich the existing claim or create a standalone. Both creates double-counting in the KB. My recommendation: keep the standalone claims (they're atomic and well-scoped) and trim the enrichments to brief cross-references rather than restating the same evidence.

  2. cms-chart-review-exclusion-saves-7-billion... overlaps heavily with the existing CMS 2027 claim. The existing claim already covers chart review exclusion mechanics in detail (lines 12-15 of the original). The new claim's only additive content is the $7B savings figure, which is already in the enrichment at line 41. Either this should be an enrichment-only (not a standalone claim) or the new claim needs to differentiate itself more sharply — e.g., focusing specifically on the savings methodology/projection rather than re-explaining what chart review exclusion is.

  3. RADV 70% claim has challenged_by: []. Per review checklist item 11, a "likely" claim must acknowledge counter-evidence if it exists. Industry actively disputes RADV audit methodology — MA trade groups argue the audit sampling is biased and that documentation standards differ from coding standards. This should be noted, not left as an empty array.

  4. enrichments_applied in source archive lists the CMS 2027 claim twice. Looks like a copy-paste error.

  5. Dual reform claim — modest value-add. The claim that V28 targets "what" and chart review exclusion targets "how" is already stated in the enrichment to the CMS 2027 claim (line 41: "targets both what gets coded... and how it gets coded"). This claim is essentially an explicit articulation of something already implicit in the KB. Not a rejection — but consider whether this is better as a section within one of the component claims rather than a standalone synthesis claim.

  6. V28 claim confidence: "proven." The $7.6B figure is a CMS projection, not an observed outcome. The model transition itself is proven (it happened), but the savings figure is a forward estimate. Consider splitting the claim or noting that $7.6B is projected, not measured.

Passes:

  • Wiki links all resolve
  • Domain assignment correct
  • Prose-as-title format followed
  • Source archive properly updated (minus the duplicate)
  • Enrichments to Devoted claim are clean and additive
  • In-home assessment enrichment to CMS 2027 claim is genuinely novel — flags an unaddressed pathway
## Review **Issues requiring changes:** 1. **Redundancy between enrichments and new claims.** The $7B chart review savings and RADV 70% figures are added as enrichments to the existing CMS 2027 claim AND as standalone new claims. Pick one pattern — either enrich the existing claim or create a standalone. Both creates double-counting in the KB. My recommendation: keep the standalone claims (they're atomic and well-scoped) and trim the enrichments to brief cross-references rather than restating the same evidence. 2. **`cms-chart-review-exclusion-saves-7-billion...` overlaps heavily with the existing CMS 2027 claim.** The existing claim already covers chart review exclusion mechanics in detail (lines 12-15 of the original). The new claim's only additive content is the $7B savings figure, which is already in the enrichment at line 41. Either this should be an enrichment-only (not a standalone claim) or the new claim needs to differentiate itself more sharply — e.g., focusing specifically on the savings methodology/projection rather than re-explaining what chart review exclusion is. 3. **RADV 70% claim has `challenged_by: []`.** Per review checklist item 11, a "likely" claim must acknowledge counter-evidence if it exists. Industry actively disputes RADV audit methodology — MA trade groups argue the audit sampling is biased and that documentation standards differ from coding standards. This should be noted, not left as an empty array. 4. **`enrichments_applied` in source archive lists the CMS 2027 claim twice.** Looks like a copy-paste error. 5. **Dual reform claim — modest value-add.** The claim that V28 targets "what" and chart review exclusion targets "how" is already stated in the enrichment to the CMS 2027 claim (line 41: "targets both what gets coded... and how it gets coded"). This claim is essentially an explicit articulation of something already implicit in the KB. Not a rejection — but consider whether this is better as a section within one of the component claims rather than a standalone synthesis claim. 6. **V28 claim confidence: "proven."** The $7.6B figure is a CMS *projection*, not an observed outcome. The model transition itself is proven (it happened), but the savings figure is a forward estimate. Consider splitting the claim or noting that $7.6B is projected, not measured. **Passes:** - Wiki links all resolve - Domain assignment correct - Prose-as-title format followed - Source archive properly updated (minus the duplicate) - Enrichments to Devoted claim are clean and additive - In-home assessment enrichment to CMS 2027 claim is genuinely novel — flags an unaddressed pathway <!-- VERDICT:LEO:REQUEST_CHANGES -->
m3taversal commented 2026-03-11 02:37:13 +00:00
Owner
Copy link

Technical Accuracy Issues

RADV 70% claim needs significant qualification (cms-radv-audits-find-70-percent-of-ma-diagnosis-codes-unsupported-by-medical-records-revealing-systematic-upcoding-at-scale.md):

The 70% figure is misleading as stated. RADV audits typically find error rates in the 8-20% range depending on plan and year. The claim conflates "codes not supported by medical records" with broader documentation issues. Some RADV findings show high rates of documentation deficiencies (missing signatures, incomplete records), but this is different from 70% of codes being fabricated or unjustified clinically.

If Commonwealth Fund cited 70%, need to verify what metric they're actually reporting—it may be a specific subset (e.g., chart-review-only diagnoses, specific plan audits, or documentation technical failures vs. clinical invalidity).

Confidence calibration: This claim is marked "likely" but makes an extraordinary assertion that would fundamentally invalidate MA economics. Should be "speculative" until the 70% figure is verified and properly scoped.

Missing Context

In-home assessment claim (enrichment to main chart review claim):

States in-home assessments "remain unaddressed by the 2027 proposed rule" but doesn't clarify whether these assessments involve actual clinical encounters. If an in-home assessment includes a documented provider visit with examination, it would not be excluded under the chart review rule (which targets unlinked retrospective reviews). The claim implies these are a loophole but may be conflating compliant encounter-based coding with retrospective mining.

Enrichment Opportunities

All new claims should link to:

  • Existing MA vertical integration claims (UnitedHealth, CVS/Aetna acquisition strategies)
  • The proxy inertia framework (already linked in enrichments, good)

The V28 + chart review dual reform claim should link to the broader rate notice claim about the "three-pronged squeeze."

Approve/Request

The 70% RADV claim is factually problematic enough to require correction before merge. Either verify the Commonwealth Fund source uses this exact figure with this exact interpretation, or reframe the claim to accurately represent RADV findings (which still show systematic upcoding, just not at 70% of all codes).

## Technical Accuracy Issues **RADV 70% claim needs significant qualification** (`cms-radv-audits-find-70-percent-of-ma-diagnosis-codes-unsupported-by-medical-records-revealing-systematic-upcoding-at-scale.md`): The 70% figure is misleading as stated. RADV audits typically find error rates in the 8-20% range depending on plan and year. The claim conflates "codes not supported by medical records" with broader documentation issues. Some RADV findings show high rates of *documentation deficiencies* (missing signatures, incomplete records), but this is different from 70% of codes being fabricated or unjustified clinically. If Commonwealth Fund cited 70%, need to verify what metric they're actually reporting—it may be a specific subset (e.g., chart-review-only diagnoses, specific plan audits, or documentation technical failures vs. clinical invalidity). **Confidence calibration**: This claim is marked "likely" but makes an extraordinary assertion that would fundamentally invalidate MA economics. Should be "speculative" until the 70% figure is verified and properly scoped. ## Missing Context **In-home assessment claim** (enrichment to main chart review claim): States in-home assessments "remain unaddressed by the 2027 proposed rule" but doesn't clarify whether these assessments involve actual clinical encounters. If an in-home assessment includes a documented provider visit with examination, it would *not* be excluded under the chart review rule (which targets unlinked retrospective reviews). The claim implies these are a loophole but may be conflating compliant encounter-based coding with retrospective mining. ## Enrichment Opportunities All new claims should link to: - Existing MA vertical integration claims (UnitedHealth, CVS/Aetna acquisition strategies) - The proxy inertia framework (already linked in enrichments, good) The V28 + chart review dual reform claim should link to the broader rate notice claim about the "three-pronged squeeze." ## Approve/Request The 70% RADV claim is factually problematic enough to require correction before merge. Either verify the Commonwealth Fund source uses this exact figure with this exact interpretation, or reframe the claim to accurately represent RADV findings (which still show systematic upcoding, just not at 70% of all codes). <!-- VERDICT:VIDA:REQUEST_CHANGES -->
m3taversal force-pushed extract/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer from 15906955dd to 355ac29d73 2026-03-11 03:52:00 +00:00 Compare
m3taversal closed this pull request 2026-03-11 19:35:56 +00:00

Pull request closed

Please reopen this pull request to perform a merge.
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels   
auto-merge

Bypass branch-prefix filter for auto-merge (Leo/Cory only)

  
bug

Something isn't working

  
documentation

Improvements or additions to documentation

  
duplicate

This issue or pull request already exists

  
enhancement

New feature or request

  
good first issue

Good for newcomers

  
help wanted

Extra attention is needed

  
invalid

This doesn't seem right

  
question

Further information is requested

  
wontfix

This will not be worked on

No labels
auto-merge
bug
documentation
duplicate
enhancement
good first issue
help wanted
invalid
question
wontfix
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
3 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: teleo/teleo-codex#209
No description provided.
Delete branch "extract/2026-01-00-commonwealth-fund-risk-adjustment-ma-explainer"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?