How SNFs actually
make money
A complete breakdown of Medicare Part A reimbursement, PDPM mechanics, and exactly where revenue leaks — written for someone who knows MDS but not the financial engine it drives.
Who Pays SNFs and How Much
SNFs get paid by four payer types. The financial stakes are wildly different depending on which one:
| Payer | Typical Daily Rate | % of Revenue | What Drives the Rate |
|---|---|---|---|
| Medicare Part A | $500 – $1,100+/day | ~30–40% | PDPM case mix — acuity of the resident |
| Medicare Advantage (MA) | $350 – $700/day | ~20–30% | Negotiated contract rates — flat or banded |
| Medicaid | $180 – $300/day | ~30–45% | State rate-setting formulas |
| Private Pay | $200 – $400+/day | ~5–15% | Facility’s published rates |
Medicare Part A is the highest-paying payer by a wide margin. A resident in a skilled nursing stay can generate 3–5x the daily revenue of a Medicaid long-term resident in the same bed. And unlike Medicaid, where the state sets the rate, Medicare Part A rates are not fixed — they’re calculated per-resident based on acuity. Your coding determines your payment.
Medicare Part A doesn’t pay a fixed rate per day. It pays a rate that reflects how sick and resource-intensive each specific resident is. That rate is determined almost entirely by how the MDS is coded. Better coding = higher rate = more revenue for the exact same care delivered.
The PDPM Payment System
PDPM stands for Patient-Driven Payment Model. CMS replaced the old RUG-IV system with it in October 2019. Under RUG-IV, therapy minutes drove payment. PDPM shifted the basis of payment to clinical complexity. That makes accurate coding more important, not less.
Under PDPM, every Medicare Part A resident is assigned a per diem rate calculated by adding up five separate payment components. Each component has its own case mix adjustment — its own multiplier based on coded clinical conditions.
Under RUG-IV, a facility could hit a high payment category by scheduling enough therapy minutes. Under PDPM, you can’t manufacture a higher rate through scheduling decisions — the rate comes from diagnosis codes and functional assessments on the MDS. The MDS coordinator directly controls the facility’s revenue on every Medicare admission.
PDPM has a rate adjustment built in for length of stay. Most components pay at full rate for days 1–20, then step down by a certain percentage after day 20. The first 20 days are where the highest rates apply. If a resident’s admission assessment is coded inaccurately, the facility loses money at its highest-rate window and never recovers it.
The 5 Payment Components
Each component is calculated separately, assigned a case mix group, and combined with a base rate. Here’s what drives each one:
Primary Dx
Therapy need
Cognitive function
Primary Dx
Cognitive impairment
Trach/vent
Special treatments
Depression/anxiety
High-cost Dx
IV meds/treatments
The PT, OT, and SLP components are driven by Section GG functional scores combined with the primary diagnosis. The NTA component is different from all of them — and it’s the one with the most money on the table.
The NTA — where the big money is
NTA stands for Non-Therapy Ancillary. It represents the cost of high-intensity clinical services that aren’t therapy — IV medications, complex wound care, dialysis, specialty drugs, high-acuity diagnoses. CMS created a comorbidity scoring table where each qualifying diagnosis or treatment adds a point value. Higher score = higher daily rate for the NTA component.
Days 1–3 pay at a 3× NTA multiplier. The NTA component pays at triple rate for days 1 through 3 of a Medicare Part A stay. Day 4 onward drops to the standard rate. A coding gap that depresses the NTA score costs the facility money at maximum multiplier — it is the single most expensive coding error a facility can make, and it happens at the most chaotic point of any admission.
If the correct NTA score for a resident is 8 but only 4 is coded, the facility isn’t just losing the difference in daily rate — it’s losing that difference at 3× on days 1 through 3. The error is magnified exactly when it hurts most.
| Diagnosis / Condition | NTA Pts | Where It Lives | Why It Gets Missed |
|---|---|---|---|
| IV medications (3+ days) | 3 | MAR, physician orders | Nursing documents it; nobody connects MAR → MDS Section O |
| Parenteral/IV feedings | 3 | Section K, nutrition orders | Flagged in nutrition but not cross-referenced to MDS |
| Septicemia | 3 | Discharge summary, H&P | Hospital codes “infection” vaguely; specific ICD not mapped |
| Tracheostomy care | 3 | Section O, nursing notes | Charted as a care task; MDS checkbox not triggered |
| Dialysis | 3 | Section O, transport orders | Known to care team; MDS section not updated |
| Pneumonia (active) | 2 | Discharge summary | Hospital documents “respiratory failure” — lower weight |
| HIV/AIDS | 2 | Face sheet, H&P | Listed as historical dx; active status not flagged |
| Multiple sclerosis | 2 | H&P, problem list | Mapped under Section I but NTA weight not recognized |
| Quadriplegia | 3 | H&P, face sheet | Obvious to nursing; sometimes coded too broadly in MDS |
| Isolation precautions | 1 | Infection control notes | Nursing knows; coordinator doesn’t always see infection log |
How MDS codes drive dollars
As an MDS coordinator, you complete the assessment. Here’s the direct wiring from each section to a payment calculation:
| MDS Section | What You’re Coding | What It Drives |
|---|---|---|
| Section A | Admission, demographics, payer | Assessment type, billing period |
| Section GG | Functional status (ADLs at admission) | PT, OT case mix group — major driver |
| Section I | Active diagnoses checked | NTA comorbidity score, SLP component, nursing |
| Section J | Health conditions (swallowing, pain, falls) | SLP case mix, nursing component |
| Section K | Swallowing/nutritional status | SLP component, parenteral feeding NTA points |
| Section O | Special treatments, procedures, programs | Nursing component, NTA (IV meds, trach, dialysis) |
Every dollar of PDPM payment flows through something you coded on that MDS. If the MDS doesn’t say it, it doesn’t get paid.
Clinical documentation → MDS coding → PDPM calculation → claim submission → Medicare payment. The only point in that chain where accuracy can be improved is the MDS coding step.
Why money gets left behind
This is not a competence problem. MDS coordinators who miss revenue aren’t bad at their jobs — they’re working in a structurally broken information environment.
- Documentation lives in multiple systems. The diagnosis is in the hospital’s discharge summary. The IV medication is in the MAR. The isolation status is in the infection control log. The MDS coordinator is expected to synthesize all of this under admission pressure — often while managing multiple concurrent assessments.
- Hospital coding doesn’t optimize for SNF payment. Hospital coders code for DRG. They use the ICD-10 code that’s specific enough to support the hospital’s claim, which is often not the most specific code available. “Respiratory failure” pays the hospital fine. “Pneumonia” is equally supported by the same chart and adds NTA points for the SNF. Nobody tells you this at discharge.
- The NTA comorbidity table isn’t part of clinical training. The NTA comorbidity weight of specific diagnoses is something you either studied in post-hire training or you didn’t. Most coordinators work from pattern recognition, not a systematic table lookup every time.
- The 5-day window creates impossible time pressure. The 5-day MDS has to be locked within the assessment reference date window. Coordinators are working against the clock, coordinating with therapy, gathering documentation, and often covering multiple assessments at once.
- No feedback loop. When a code gets missed, nobody tells you. The claim goes out, the payment comes back, and the gap between what was paid and what could have been paid is invisible unless someone does a retrospective audit.
The ICD-10 Specificity Problem
ICD-10 is a hierarchical coding system. Hospital coders optimize for DRG, not for SNF PDPM. They’ll document the code that’s defensible for their claim — which is often not the most specific code available. CMS’s NTA comorbidity table doesn’t reward all codes in a family equally. It rewards specificity.
| Hospital Code | More Specific Code | Clinical Basis | Payment Impact |
|---|---|---|---|
| Altered mental status (R41.3) | Maj. neurocog disorder, vascular (F01.51) | Same chart, more specific | SLP flag + NTA points |
| Acute respiratory failure (J96.00) | Pneumonia (J18.9 or specific pathogen) | Often both present and documented | NTA comorbidity points |
| Sepsis (A41.9) | Septicemia — specific organism (A41.01–A41.89) | Blood culture in chart | 3 NTA points |
| Chronic kidney disease unspec. (N18.9) | CKD Stage 4 or 5 (N18.4/N18.5) | GFR documented in labs | NTA comorbidity points |
| Pressure ulcer (L89.90) | Stage-specific pressure ulcer code | Wound assessment in chart | NTA + QM flags |
You can only use a more specific code if the documentation supports it. This isn’t upcoding — it’s accurate coding. The hospital’s code was defensible; a more specific code is equally defensible and pays better. A coding consultant doing this manually charges $150–250/hour. The right workflow does it in seconds on every admission.
What the tool does to fix it
Every revenue leak described above has a corresponding fix. Here’s the direct mapping:
Diagnoses buried in discharge summary, not systematically cross-referenced to MDS Section I
Ingests discharge summary, parses diagnoses, maps each one to MDS checkboxes and NTA comorbidity table automatically
IV medications documented in MAR but Section O NTA checkboxes not triggered
Cross-references MAR/physician orders against Section O items and surfaces uncoded procedures with NTA point values
Hospital ICD-10 codes are DRG-optimized, not SNF-optimized for PDPM
Suggests more specific ICD-10 codes equally supported by chart documentation, with higher PDPM weights
No real-time dollar impact visible for each coding decision
Calculates revenue delta per uncoded item in $/day and projected LOS total — makes the gap undeniable to administrators
No documentation checklist connecting each MDS code to the chart evidence required by F641
Generates documentation requirements per flagged item: “To support IV meds coding — physician order + 3-day MAR entries in look-back period”
The math: a real example
Meet Mrs. R. She’s a 78-year-old admitted after a UTI with sepsis. Her MDS 5-day is due in 3 days. Here’s what the hospital discharge summary documents vs. what gets coded:
| Clinical Finding | In the Chart? | Coded on MDS? | NTA Points | 3× Days 1–3 |
|---|---|---|---|---|
| Septicemia (E. coli) | ✓ Discharge summary | ✗ Not coded | 3 pts | 9 pts-days |
| IV antibiotics × 5 days | ✓ MAR + orders | ✗ Not in Section O | 3 pts | 9 pts-days |
| Isolation precautions | ✓ Infection control | ✗ Missed | 1 pt | 3 pts-days |
| CKD unspecified (GFR 35) | ✓ Labs documented | Wrong code | 1 pt diff. | 3 pts-days |
Total missed NTA points: 8 points. At a base NTA rate of approximately $50/point/day, factoring the 3× multiplier for days 1–3:
Days 1–3: 8 missed points × $50/pt × 3× multiplier = $1,200 lost over 3 days
Days 4–60: 8 missed points × $50/pt × 57 days = $22,800 lost over remainder of stay
Total for this one resident: ~$24,000 left on the table.
This facility does 12 admissions per month. Even if only half have coding gaps half this size, that’s roughly $72,000/month — $864,000/year — walking out the door.
Compliance = revenue protection
Revenue capture without documentation is fraud. Every PDPM code that drives payment needs to be defensible in the clinical record. The F-tags you’re already working with are the enforcement mechanism:
| F-Tag | What It Requires | Revenue Connection |
|---|---|---|
F641 | MDS must accurately reflect the resident’s status as documented in the clinical record | Every PDPM-driving code must have chart support — the compliance requirement and the revenue requirement are identical |
F636 | Comprehensive assessment completed timely | Late 5-day MDS means delayed payment start and potential payment reduction |
F637/638 | Significant change assessments triggered and completed | Missed SCAs mean payment rate doesn’t adjust when clinical complexity increases |
A MAC auditor and your PDPM billing optimizer are asking the exact same question: “Is there clinical documentation in the look-back period that supports this MDS code?” There is no version of this where compliance and revenue optimization are in conflict. They’re the same activity.
You already know how to do the assessment.
Now close the gap.
The tool gives you a systematic cross-reference between what’s documented in the chart and what’s coded on the assessment — with the dollar impact attached.