Phase 2 · Methodology Card · v1.1
Methodology Card
Produce chart-backed, defensible evidence of earned-but-unbilled / under-documented revenue and documentation risk. Findings must be reproducible and usable by operators, lenders, investors, REITs, and PE.
Single source of truth
Intelligence Methodology Card
This card governs every Intelligence project. Print it. Follow it. Sign against it at QC. Any deviation must be documented and approved before delivery.
DRL Holdings Intelligence · Methodology Card · v1.1
Purpose
Produce chart-backed, defensible evidence of earned-but-unbilled / under-documented revenue and documentation risk. Findings must be reproducible and usable by operators, lenders, investors, REITs, and PE.
Core process (mandatory sequence)
| Step | Action | Output |
|---|---|---|
| 1. Intake | Confirm BAA is signed. Log facility, EMR, date range, sample size, and project ID. | Project Intake Log |
| 2. Extraction | Pull only what already exists in the chart: diagnoses, notes, procedures, MDS/PDPM elements, and billed codes (if provided). Map to standard template. | Standardized Data File |
| 3. Dual Review | Clinical review + Coding/revenue review. Only count opportunities that are already documented on the chart. | Annotated Findings Log |
| 4. Quantification | Calculate sample-level gap dollars and annualized estimate (with confidence range). Separate: Opportunity $ vs. Audit Exposure $. | Gap Calculation Sheet |
| 5. Scoring | Assign two scores (0–100): Revenue Opportunity Score and Documentation Risk Score. | Score Summary |
| 6. Packaging | Translate the same findings into the audience-specific format (Operator / Lender / Investor / REIT / PE). | Final Report + Evidence Package |
| 7. QC & Delivery | Second review for methodology adherence. Deliver report + source evidence. | Signed-off Deliverable |
Non-negotiable rules
- Evidence only — nothing is invented or assumed.
- Flat-fee pricing only — no percentage of recovery.
- BAA required before any PHI is accessed.
- Every finding must be traceable to a specific chart note or code.
- Process must be reproducible by another qualified reviewer.
Scoring definitions (quick reference)
- Revenue Opportunity Score (0–100): Magnitude and confidence of supported revenue currently left uncaptured.
- Documentation Risk Score (0–100): Strength of documentation relative to what was billed / potential audit exposure.
Version control: This card is the single source of truth for every Intelligence project. Any deviation must be documented and approved before delivery.
QC sign-off: ______________________ Date: ____________ · Deviation log attached: Yes / No
Mandatory sequence
Core process — step, action, output
Same seven steps on every engagement. Each step produces a named artifact so another reviewer can reconstruct the work.
- 1. Intake → Project Intake Log Confirm BAA is signed. Log facility, EMR, date range, sample size, and project ID. No PHI until BAA is in place.
- 2. Extraction → Standardized Data File Pull only what already exists in the chart: diagnoses, notes, procedures, MDS/PDPM elements, and billed codes (if provided). Map to the standard template.
- 3. Dual Review → Annotated Findings Log Clinical review + coding/revenue review. Only count opportunities that are already documented on the chart. Nothing invented.
- 4. Quantification → Gap Calculation Sheet Sample-level gap dollars and annualized estimate with confidence range. Keep Opportunity $ and Audit Exposure $ separate.
- 5. Scoring → Score Summary Assign Revenue Opportunity Score and Documentation Risk Score (each 0–100) using the rubric.
- 6. Packaging → Final Report + Evidence Package Translate the same findings into the audience-specific format: Operator, Lender, Investor, REIT, or PE. Every finding cites chart evidence.
- 7. QC & Delivery → Signed-off Deliverable Second review for methodology adherence. Deliver report + source evidence. Optional walkthrough call.
0–100 scores
Scoring rubric
Revenue Opportunity Score: magnitude and confidence of supported revenue currently left uncaptured. Documentation Risk Score: strength of documentation relative to what was billed / potential audit exposure. Both are derived from quantified gaps, then annualized, using the same bands on every project.
Revenue Opportunity Score
Magnitude and confidence of supported revenue currently left uncaptured. Primary input: annualized earned-but-unbilled estimate ÷ current Part A (or in-scope) revenue, mid-point of the confidence range.
| Annualized opportunity % of revenue | Base score |
|---|---|
| < 1% | 0–15 |
| 1% – < 3% | 16–35 |
| 3% – < 6% | 36–55 |
| 6% – < 10% | 56–75 |
| 10% – < 15% | 76–90 |
| ≥ 15% | 91–100 |
- Confidence narrow (±25% or better): use band mid-point as scored.
- Confidence wide (±25% to ±50%): subtract up to 8 points.
- Concentration risk: if >50% of opportunity sits in one diagnosis family, subtract up to 5 (execution risk for the buyer).
- Floor: unsupported / invented findings score zero contribution. Evidence-only.
Documentation Risk Score
Strength of documentation relative to what was billed / potential audit exposure. Primary input: annualized audit-exposure estimate ÷ current Part A (or in-scope) revenue, mid-point of the confidence range.
| Annualized exposure % of revenue | Base score |
|---|---|
| < 1% | 0–15 |
| 1% – < 3% | 16–35 |
| 3% – < 6% | 36–55 |
| 6% – < 10% | 56–75 |
| 10% – < 15% | 76–90 |
| ≥ 15% | 91–100 |
- +5 to +12: contradictory notes vs. billed codes (high take-back severity).
- +3 to +8: missing required elements for billed PDPM/HIPPS drivers.
- Cap at 100. Narrative must name the top drivers with chart cites.
How scores appear in delivery
- Dashboard pair Both scores on one visual — opportunity left vs. audit heat — so capital and operators see the same facts.
- Narrative Written memo explains the mid-point dollars, confidence range, top three drivers, and what would move each score.
- Not a credit rating Scores are documentation intelligence, not a lending decision, investment recommendation, or coding opinion for billing.
v1 · Comparable projects
Standardized extraction template
One row per chart (or per stay in the sample). Every project maps into these fields before dual-layer review. Download the CSV starter for intake workbooks.
| Field | What to capture | Source |
|---|---|---|
| project_id | Unique engagement ID | Intake log |
| facility_id / site_label | De-identified facility key (PHI-safe label) | Intake |
| chart_id / stay_id | Internal sample identifier (not MRN in shared workbooks) | Export map |
| emr_source | EHR / export format (PDF, CCD, HL7, FHIR, text) | Intake |
| date_range | Stay or sample window | Export |
| diagnoses_icd10 | Listed / coded diagnoses on file | Chart / claim |
| procedures_treatments | Procedures and treatments documented | Notes / orders |
| note_types_present | Physician / nursing / therapy / MAR flags | Export inventory |
| mds_pdpm_elements | MDS / PDPM-related documentation when applicable | MDS / chart |
| billed_codes_claims | Existing billed codes / claims if provided | Claims file |
| supported_uncaptured | Diagnoses/services supported in notes, not on claim | Dual-layer review |
| weak_or_contradictory | Docs weak/contradictory vs. billed items | Dual-layer review |
| evidence_cites | Note type, date, clinician / author string for each finding | Chart |
| opp_dollars_sample | Earned-but-unbilled $ for this row | Quantification |
| risk_dollars_sample | Audit exposure $ for this row | Quantification |
| reviewer / qc_status | Primary reviewer + QC pass/fail | QC |
Minimum viable sample
40 charts as the default
Decision: minimum viable sample for a single-facility estimate is 40 Medicare Part A charts drawn from the agreed date range, with an acceptable floor of 30 and a preferred band of 40–50.
Below 30, absolute dollars for the sample may still be useful for coaching, but annualized facility estimates must be labeled exploratory and confidence ranges widen materially. Above 50 improves precision when case-mix is highly heterogeneous or the buyer needs tighter bands for capital decisions.
- 30 charts Floor for directional work. Annualized estimate requires wide confidence language.
- 40 charts Default MVP — reliable enough for operator scorecards and most lender / PE memos.
- 40–50 charts Preferred when mix varies by HIPPS / clinical category or when REIT/PE diligence needs tighter ranges.
- Multi-site / portfolio Stratify by operator or building; target ≥30 per stratum or a documented proportional sample with explicit coverage notes.
Non-negotiable
Rules that never bend
- Evidence only Nothing is invented or assumed. Every finding must be traceable to a specific chart note or code.
- Flat-fee pricing only No percentage of recovery.
- BAA required No PHI is accessed until a Business Associate Agreement is in place.
- Reproducible Another qualified reviewer following the same steps should reach substantially the same conclusions.
- Version control This card is the single source of truth. Any deviation must be documented and approved before delivery.