13.3 All-Patient Refined DRG (APR-DRG) System

All-Patient Refined DRG (APR-DRG) System

AHIMA CCS Exam Focus: While CMS utilizes MS-DRGs for Medicare inpatient reimbursement, state Medicaid programs, commercial payers, pediatric health systems, and quality rating organizations rely heavily on the Solventum All-Patient Refined Diagnosis Related Group (APR-DRG) system. Candidates must understand the dual-subclass structure (Severity of Illness [SOI] vs. Risk of Mortality [ROM]), how APR-DRGs evaluate cumulative multi-organ secondary diagnosis interactions (contrasting with MS-DRG's single-highest CC rule), and the application of APR-DRGs in hospital quality profiling and expected mortality modeling.

---

1. Evolution and Purpose of the APR-DRG System

The original Medicare DRG system (and its modern MS-DRG successor) was designed specifically for the Medicare population—individuals aged 65 and older or those with permanent disabilities and end-stage renal disease (ESRD). As a result, MS-DRGs exhibit substantial structural limitations when applied to broader patient demographics:

• Homogeneous Pediatric & Neonatal Grouping: MS-DRGs lack clinical granularity for low-birthweight neonates, congenital pediatric anomalies, and pediatric intensive care admissions.
• Obstetric Limitations: MS-DRGs categorize most maternity and delivery admissions into broad, flat groups with minimal severity distinction.
• Single-Payer Focus: MS-DRGs prioritize financial resource consumption rather than clinical morbidity and mortality profiling across diverse age spectra.

To resolve these shortcomings, Solventum (formerly 3M Health Information Systems), in collaboration with the National Association of Children's Hospitals and Related Institutions (NACHRI), developed the All-Patient Refined DRG (APR-DRG) system. APR-DRGs are a widely used licensed inpatient methodology in settings such as:

1. Some state Medicaid programs and other public payers
2. Some commercial managed-care plans
3. Organizations serving pediatric, obstetric, and other all-patient populations
4. Quality and outcomes analyses that license the applicable grouper and methodology. Payer adoption and software versions vary, so verify the contract and current grouper rather than assuming universal use.

                            MS-DRG vs. APR-DRG Scope
  ┌─────────────────────────────────────────┬─────────────────────────────────────────┐
  │           CMS MS-DRG SYSTEM             │            Solventum APR-DRG SYSTEM            │
  ├─────────────────────────────────────────┼─────────────────────────────────────────┤
  │ • Primary User: Medicare Part A         │ • Primary Users: Medicaid, Commercial,  │
  │ • Population: Elderly (65+), Disabled   │   Pediatric, Quality Rating Agencies    │
  │ • Primary Focus: Financial Resource     │ • Population: All patient demographics  │
  │   Consumption & Bed-Day Costs           │   (Neonates, Pediatrics, Adults, OB)    │
  │ • Structure: ~760 MS-DRGs with 1–3      │ • Primary Focus: Clinical Morbidity,    │
  │   severity splits per family            │   Mortality Risk, and Resource Use      │
  │ • Secondary Logic: Single-highest       │ • Structure: 332 base groups × 4 SOI   ×   │
  │   CC/MCC threshold                      │   each case also has one of 4 ROM levels  │
  └─────────────────────────────────────────┴─────────────────────────────────────────┘

---

2. The Dual-Subclass Architecture: SOI vs. ROM

The foundational innovation of the APR-DRG system is its dual-subclass architecture. In APR-DRGs, every inpatient encounter is assigned to a base APR-DRG (e.g., Base APR-DRG 139 Pneumonia) and then independently assigned two distinct 4-level subclass ratings:

                          The APR-DRG 4-Level Subclass Model
  ┌───────────────────────────────────┐   ┌───────────────────────────────────┐
  │    Severity of Illness (SOI)      │   │      Risk of Mortality (ROM)      │
  ├───────────────────────────────────┤   ├───────────────────────────────────┤
  │ Subclass 1: Minor                 │   │ Subclass 1: Minor                 │
  │ Subclass 2: Moderate              │   │ Subclass 2: Moderate              │
  │ Subclass 3: Major                 │   │ Subclass 3: Major                 │
  │ Subclass 4: Extreme               │   │ Subclass 4: Extreme               │
  └───────────────────────────────────┘   └───────────────────────────────────┘

Severity of Illness (SOI) vs. Risk of Mortality (ROM) Defined:

• Severity of Illness (SOI):

  • Definition: The extent of physiological decompensation, organ system loss of function, and overall clinical complexity.
  • Operational Application: Classifies physiologic decompensation or organ-system loss of function. Adopting payers may associate SOI subclasses and APR-DRGs with payment weights or resource analyses, but the effect is contract- and version-specific.

• Risk of Mortality (ROM):

  • Definition: The clinical likelihood of in-hospital death for the patient.
  • Operational Application: Used for risk adjustment in quality reporting, hospital-wide mortality benchmarking, surgeon-specific mortality profiling, and calculating Observed-to-Expected (O/E) Mortality Ratios.

Clinical Divergence Between SOI and ROM

A critical concept on the CCS examination is recognizing that SOI and ROM can diverge. A resource-intensive case does not automatically receive the highest mortality subclass, and a short stay does not automatically imply low mortality risk. Age, diagnoses, procedures, interactions, and the complete coded record matter. Exact subclass examples must be run through the applicable licensed grouper rather than assigned from clinical intuition alone.

---

3. APR-DRG Grouping Logic: Cumulative Multi-Organ Interactions

MS-DRG severity splits generally use the highest applicable CC/MCC tier after exclusions and the rest of the grouper logic. APR-DRG instead assigns both SOI and ROM subclasses through licensed, version-specific logic that considers the complete coded case.

                       Secondary Diagnosis Processing Logic
  ┌─────────────────────────────────────────────────────────────────────────────┐
  │ CMS MS-DRG Logic (Threshold Model):                                         │
  │ [Base DRG] + [Highest applicable severity tier after exclusions/rules]      │
  │ Additional diagnoses remain reportable and can affect other grouper logic.   │
  ├─────────────────────────────────────────────────────────────────────────────┤
  │ Solventum APR-DRG Logic (Cumulative Interaction Model):                            │
  │ [Base APR-DRG] + [Dx 1 (Moderate)] + [Dx 2 (Moderate)] + [Dx 3 (Moderate)]   │
  │                      │                                                      │
  │                      ▼ (Licensed, version-specific interaction logic)        │
  │ [Result: Grouper calculates final SOI and ROM subclasses]                    │
  └─────────────────────────────────────────────────────────────────────────────┘

How APR-DRG Calculates SOI and ROM Subclasses:

The licensed grouper evaluates the principal diagnosis, reportable secondary diagnoses, procedures, age and other patient attributes, then applies version-specific interactions and reassessment logic to calculate SOI and ROM. A diagnosis may increase, decrease, or leave a subclass unchanged depending on the entire record. Coders should report every supported condition that meets reporting rules, then use the applicable licensed grouper rather than attempting to derive an exact subclass manually.

---

4. MS-DRG vs. APR-DRG Comprehensive Comparison Matrix

The following matrix outlines the pivotal structural, operational, and regulatory distinctions between the two grouping systems:

Clinical Documentation Integrity (CDI) & Quality Use

Complete, accurate reporting matters under APR-DRGs because multiple valid secondary diagnoses and their interactions may affect SOI or ROM under version-specific logic. Report every condition that meets official reporting rules, but do not add diagnoses merely to influence a subclass.

APR-DRG SOI and ROM may be used in payment, internal analytics, or risk-adjusted quality measurement depending on the payer and organization. The effect of one diagnosis cannot be predicted reliably without the licensed grouper version and the full coded record.