PDPM admissions guide

PDPM major case-mix classification at the SNF admission decision point.

PDPM does not start at the 5-day PPS assessment. Most of the inputs that drive PT, OT, SLP, Nursing, and NTA classification can already be read out of the hospital referral packet, which is exactly where admissions teams need them.

Why this matters at intake

Why admissions teams should read PDPM signals before the bed is held.

Under the Patient Driven Payment Model, SNF Part A per-diem rates are case-mix adjusted across five components: Physical Therapy (PT), Occupational Therapy (OT), Speech-Language Pathology (SLP), Nursing, and Non-Therapy Ancillaries (NTA). A non-case-mix component is also part of the per-diem. Each component classifies the resident independently using clinical category, MDS items, function score, and comorbidity flags defined in CMS guidance.

The formal PDPM classification is locked in at the 5-day PPS assessment, but most of the inputs that drive that classification are already in the hospital discharge packet on the day of referral: principal diagnosis, ICD-10 comorbidities, IV medications, swallowing and cognition signals, prior level of function, and functional-status documentation. Reading those signals before acceptance helps admissions teams flag missing documentation, ask the hospital for clarification, and plan staffing or NTA exposure earlier in the process.

The five case-mix components

How PT, OT, SLP, Nursing, and NTA each classify the same resident.

Each component runs its own classification logic against the same clinical record. The four headings below describe the inputs admissions teams can usually see in a referral packet, framed as questions to verify rather than rates to predict.

PT and OT components

PDPM groups PT and OT classification around the resident's clinical category (derived from the primary reason for SNF care) and a function score built from MDS Section GG self-care and mobility items. Some clinical categories collapse into combined PT/OT groups under CMS rules.

SLP component

SLP classification considers presence of an acute neurologic condition, a qualifying SLP-related comorbidity, cognitive impairment, swallowing disorder, and mechanically altered diet. A referral packet that does not document swallowing status or cognition leaves the SLP picture incomplete.

Nursing component

The Nursing component uses a clinical category-driven logic similar to the prior RUG-IV nursing groups, plus function score, depression indicator, and restorative nursing services. Conditions such as tracheostomy, isolation for active infection, and parenteral or IV feeding can drive significantly higher nursing classifications.

NTA component

Non-Therapy Ancillary classification is built from a comorbidity score. Each qualifying condition or treatment, including HIV/AIDS, parenteral/IV feeding, certain endocrine and respiratory conditions, and select medication regimens, contributes weighted points. NTA also uses variable per-diem adjustment, paying more in the first three days of stay than later.

Variable per-diem adjustment

PT and OT per-diem amounts decline over the course of the stay under CMS-defined adjustment factors, and NTA pays a higher rate early in the stay. Length-of-stay expectations therefore interact with the case-mix picture, not just total revenue.

Non-case-mix component

Every PDPM stay also includes a non-case-mix component covering room-and-board type costs. It is not adjusted by patient characteristics, so it does not change the admissions math but should be remembered when describing PDPM publicly.

Looks profitable, classifies high-cost

Why some referrals look attractive on the packet and classify high-cost in practice.

A short anticipated length of stay or a familiar primary diagnosis does not, on its own, predict PDPM classification or NTA exposure. The five components classify independently, and each can be driven by something other than the headline reason for admission. Educational review frameworks treat case-mix questions and clinical-capability questions separately for that reason.

Unverified IV antibiotic regimens

An IV antibiotic referenced in the discharge summary but not confirmed in the MAR may still trigger a high NTA classification when it is later documented. The drug, the route, and the planned duration belong in the pre-acceptance question set, not in the post-admission surprise pile.

Cognition and swallowing under-documented

An SLP comorbidity, swallowing disorder, or mechanically altered diet noted only inside therapy progress notes can flip the SLP group at the 5-day assessment. Admissions teams reading only the discharge summary may not see this until staffing is already committed.

Comorbidity stack on the problem list

NTA points accumulate. A resident with several individually moderate comorbidities (endocrine, respiratory, hematologic, infectious) can land in a higher NTA group than a single high-acuity diagnosis. The problem list deserves a closer read than it usually gets at intake.

PDPM-aware admissions review questions Use with source-document verification
No PHI
  • Is the primary reason for SNF care clearly stated and supported by the discharge summary?
  • Are Section GG-style functional-status items documented well enough to support a function score?
  • Does the packet show cognition status, swallowing screen, and current diet texture?
  • Are IV antibiotics, parenteral feeding, isolation status, and tracheostomy care confirmed in the MAR and nursing notes?
  • Has the comorbidity problem list been compared against the ICD-10 codes that drive PDPM groupings?
  • Are clinical-fit questions kept separate from PDPM planning context so decisions are not conflated?
Where AdmitScore fits

AdmitScore keeps PDPM signals inside admissions review, not coding work.

AdmitScore™ does not replace the MDS coordinator and is not a billing or coding tool. It treats PDPM-related signals as part of AI admissions analysis for SNFs: referral packet review, payer authorization readiness, documentation gaps, care-fit questions, and admin-only financial planning context. Final classification still belongs to the clinical and MDS team at the 5-day PPS assessment.

  • Admissions-level scope: review questions, not reimbursement guarantees.
  • Clinical-safe output: staff verify everything before relying on it.
  • Operator language: separates clinical fit from PDPM planning context for admins, DONs, and admissions coordinators.
Source orientation

Use CMS materials for the rules; use AdmitScore for the admissions workflow.

PDPM logic is defined and maintained by CMS. The summaries above are educational and should not replace the official guidance for coding, MDS assessment, or reimbursement decisions. Facility policies and qualified clinical staff make final determinations.

CMS PDPM resources

CMS publishes the PDPM clinical categories, function score logic, comorbidity lists, and variable per-diem adjustment factors. Admissions teams should treat those materials as authoritative.

Open CMS PDPM resources

CMS RAI manual

Section GG self-care and mobility items, swallowing and cognition coding, and other MDS inputs that feed PDPM all trace back to the RAI manual. Use it for coding rules, not marketing summaries.

Open CMS RAI manual resources

AdmitScore pilot

Pilot AdmitScore with synthetic-first review, PHI-free public forms, and secure workflow scoping before any live referral packet process touches it.

Request a pilot conversation
FAQ

Common PDPM admissions questions.

What are the five PDPM case-mix components?

Under the Patient Driven Payment Model, SNF Part A per-diem payment is built from five case-mix-adjusted components: Physical Therapy (PT), Occupational Therapy (OT), Speech-Language Pathology (SLP), Nursing, and Non-Therapy Ancillaries (NTA). Each classifies the resident independently using clinical category, MDS items, function score, comorbidity flags, and other inputs defined by CMS. A non-case-mix component is also part of the per-diem.

Why does PDPM matter at the admission decision point and not only at the 5-day PPS assessment?

PDPM classifications are formally set at the 5-day PPS assessment, but admissions teams can already read most of the inputs from a hospital referral packet: principal diagnosis, comorbidities, IV medications, swallowing and cognition signals, and functional-status documentation. Reading those signals before acceptance helps SNF teams flag missing documentation, ask the hospital for clarification, and plan staffing or NTA cost exposure earlier.

Can a referral look profitable on paper and still classify high-cost under PDPM?

Yes. A short anticipated length of stay or a familiar clinical category does not, on its own, predict the case-mix classification or the NTA exposure. An unverified IV antibiotic regimen, an HIV diagnosis, a tracheostomy, parenteral nutrition, or a high comorbidity burden can drive Nursing and NTA classifications well above what a quick scan of the packet suggests.