PDPM Optimization: Maximizing Reimbursement Without Gaming the System

The PDPM Opportunity Most Facilities Are Missing

When CMS replaced the Resource Utilization Group (RUG) model with the Patient-Driven Payment Model in October 2019, the intention was clear: shift Medicare reimbursement for skilled nursing from therapy volume to patient characteristics. PDPM bases payment on the clinical complexity, functional status, and service needs of each resident rather than on the number of therapy minutes delivered. In theory, this should reward facilities that accurately document the acuity of the residents they serve.

In practice, many skilled nursing facilities are leaving significant revenue on the table, not because they’re providing inadequate care, but because their documentation and coding practices haven’t fully adapted to the PDPM framework. The gap between the care being delivered and the reimbursement being captured is one of the most common and costly operational inefficiencies in post-acute care today.

This isn’t about gaming the system. It’s about accurately reflecting the clinical reality of your resident population in the documentation that drives payment classification. When a facility under-codes because its documentation is incomplete or its MDS assessments don’t capture the full clinical picture, it isn’t being conservative. It’s being inaccurate. And inaccuracy in either direction is a compliance problem.

Understanding the Five PDPM Payment Components

PDPM calculates a per-diem rate for each Medicare Part A skilled nursing resident based on five distinct payment components. Each component is classified independently, and the sum of the five component rates determines the total per-diem reimbursement. Understanding how each component is driven by documentation and coding is essential for optimization.

1. Physical Therapy (PT)

The PT component is driven by the clinical category (derived from the primary reason for the SNF stay), Section GG functional scores, and the presence of certain comorbidities. Unlike the old RUG system, the PT rate is not based on the number of therapy minutes delivered. Instead, it reflects the expected therapy resource needs based on the resident’s clinical profile. Facilities that accurately document functional limitations and relevant comorbidities ensure appropriate PT classification.

2. Occupational Therapy (OT)

Similar to the PT component, OT classification is based on clinical category, Section GG functional scores, and comorbidity presence. The OT and PT components can classify at different levels for the same resident if the functional assessment captures distinctions in the types of functional limitations present. Complete Section GG documentation is critical for both therapy components.

3. Speech-Language Pathology (SLP)

The SLP component has the widest reimbursement range of any PDPM component, making it the area with the greatest financial impact from accurate versus inaccurate coding. SLP classification is driven by the presence of SLP-related comorbidities, cognitive impairment indicators (including BIMS scores), swallowing disorders, and mechanically altered diet requirements. Many facilities under-code the SLP component because they fail to document the full range of cognitive and communication deficits that their residents present.

4. Nursing

The nursing component reflects the expected nursing resource intensity based on the resident’s clinical condition. It is driven by the clinical category, specific comorbidities documented as active during the assessment period, the use of certain services and treatments (such as IV medications, ventilator/respirator, and tracheostomy care), and the presence of depression indicators. HIV/AIDS diagnosis also affects nursing classification. The nursing component is where diagnosis documentation accuracy has the most direct financial impact.

5. Non-Therapy Ancillary (NTA)

The NTA component covers non-therapy clinical services such as medications, lab work, and medical supplies. It is driven by specific comorbidities and conditions documented in the MDS, with higher-cost conditions (such as IV medications, dialysis, and certain wound care protocols) driving higher NTA classification. Like the nursing component, NTA classification depends entirely on complete and accurate diagnosis and treatment documentation.

Where Facilities Commonly Under-Code

After analyzing PDPM coding patterns across hundreds of skilled nursing facilities, several consistent areas of under-coding emerge. These represent not just revenue opportunities but documentation accuracy gaps that should concern clinical and compliance leadership equally.

Section GG functional scoring

Section GG of the MDS captures self-care and mobility functional abilities using a performance-based scoring system. The scores directly drive the PT and OT components of PDPM. Under-coding in Section GG is pervasive, and it typically results from two factors: nursing staff who are not adequately trained on the GG scoring methodology, and documentation that captures functional status in narrative form but does not translate that narrative into the standardized scoring framework that GG requires.

The distinction between what a resident can do and what a resident does do is particularly important. Section GG captures actual performance during the assessment period, not capability. A resident who could theoretically ambulate 50 feet but consistently requires extensive assistance to do so during the assessment window should be coded based on the assistance actually provided. When nursing staff code based on potential rather than performance, the functional scores overstate independence and under-classify the resident’s therapy needs.

SLP-related comorbidities and cognitive indicators

The SLP component has the largest reimbursement differential between high and low classification levels. Yet it is consistently the most under-coded component because the conditions that drive SLP classification (cognitive impairment, aphasia, apraxia, dysphagia, and voice/communication disorders) are often documented in therapy notes but not consistently captured in the MDS assessment items that drive classification.

A resident with moderate cognitive impairment who requires cueing and supervision for safe swallowing may justify a significantly higher SLP classification, but only if the BIMS scores, swallowing assessments, and cognitive indicators are all documented and coded accurately in the MDS. Incomplete cognitive testing or undocumented swallowing assessments leave SLP revenue on the table.

Active diagnosis capture

PDPM nursing and NTA components are driven by the presence of specific diagnoses coded as active during the assessment period. “Active” means the condition affects the resident’s care, treatment, or monitoring. It does not require active treatment. A resolved fracture that still requires pain management, a history of falls that necessitates ongoing assessment, or a chronic condition that requires medication monitoring all qualify as active conditions for PDPM purposes.

Facilities that apply a narrow definition of “active” and code only conditions currently being actively treated with new interventions systematically under-capture the comorbidity burden that drives nursing and NTA classification.

AI-Enabled Coding Accuracy

The fundamental challenge of PDPM accuracy is that the people documenting clinical care (nurses, therapists, physicians) are not the same people coding the MDS, and the information needed for accurate coding is scattered across multiple documentation streams. An AI-enabled EHR addresses this challenge by creating a continuous bridge between clinical documentation and MDS coding.

• Real-time documentation-to-MDS mapping: As clinicians document care, the system identifies information that is relevant to MDS items and flags opportunities where additional documentation would support more accurate coding. If a nurse documents that a resident required extensive assistance with bed mobility, the system maps that observation to the relevant Section GG item and alerts the MDS coordinator.
• Comorbidity capture alerts: When clinical documentation references conditions that are listed on the PDPM comorbidity tables but have not been coded as active diagnoses in the MDS, the system generates an alert. This prevents the common scenario where a resident’s medical complexity is fully documented in nursing and therapy notes but not reflected in the coding that drives reimbursement.
• Section GG consistency validation: The system cross-references functional status documentation from nursing, therapy, and admission assessments to identify inconsistencies in Section GG scoring. If therapy documentation indicates a resident requires moderate assistance for transfers but the GG coding reflects supervision only, the system flags the discrepancy for review.
• SLP classification optimization: Given the significant financial impact of SLP classification, the system specifically tracks cognitive assessment completion, swallowing evaluations, and communication disorder documentation to ensure that all SLP-relevant information is captured before the assessment reference date.

“The goal is not to code higher. The goal is to code accurately. In our experience, accurate coding almost always results in higher reimbursement than the conservative, incomplete coding that most facilities default to, simply because the documentation gaps that cause under-coding are so pervasive.”

Section GG: The Most Impactful Area for Improvement

If a facility can only focus on one area of PDPM optimization, Section GG functional scoring should be that area. GG scores affect both the PT and OT components of PDPM, making it the single assessment area with the broadest reimbursement impact. And it is also the area where the gap between actual resident function and documented function is consistently widest.

Effective Section GG optimization requires:

• Standardized training: All nursing staff who document functional status must understand the GG scoring criteria, including the distinction between performance and capability, the definitions of each assistance level, and how to document concurrent activities that affect scoring.
• Observation-based documentation: GG scores should be based on direct observation during the assessment period, not on assumptions or historical performance. Staff should document specific observations that support each score, creating an audit trail that validates the coding.
• Interdisciplinary consistency: When nursing, therapy, and CNA documentation all describe the same resident’s functional status, the descriptions should be consistent. Discrepancies between disciplines create audit risk and often indicate that at least one assessment is inaccurate.
• Admission window focus: The functional assessment conducted during the first few days of the stay is the most critical for PDPM classification. Facilities should prioritize comprehensive functional status documentation during the admission window, when the resident’s level of assistance need is typically highest and the PDPM impact is greatest.

The Ethical Compliance-Revenue Balance

Any discussion of PDPM optimization must address the compliance dimension directly. The line between optimization and manipulation is clear in principle but can blur in practice. CMS actively monitors coding patterns and investigates facilities that show classification distributions inconsistent with their resident populations. The consequences of upcoding (intentionally classifying residents at higher payment levels than their clinical characteristics justify) include recovery demands, civil monetary penalties, exclusion from federal healthcare programs, and in extreme cases, criminal prosecution under the False Claims Act.

The ethical framework for PDPM optimization rests on a straightforward principle: code to the clinical reality, completely and accurately. This means:

• Never code a condition as active that does not genuinely affect the resident’s care, treatment, or monitoring
• Never inflate functional limitation scores beyond what direct observation supports
• Never document clinical findings that did not actually occur
• Always ensure that every coded item has a corresponding clinical documentation source
• Maintain an audit trail that connects coding decisions to clinical evidence

Within this framework, optimization is not only ethical. It is an obligation. Facilities that under-code because of documentation gaps are not being conservative. They are being inaccurate, and they are failing to capture the reimbursement that their residents’ genuine clinical complexity justifies. CMS designed PDPM to reimburse facilities based on the acuity of their residents. Accurate coding fulfills that design intent. Inaccurate coding, in either direction, undermines it.

Building a Sustainable PDPM Strategy

PDPM optimization is not a one-time project. It is an ongoing operational discipline that requires sustained attention to documentation quality, coding accuracy, and interdisciplinary collaboration. Facilities that treat it as a quarterly chart review exercise will see incremental improvements. Facilities that embed it into daily clinical workflows will see transformative results.

The most effective approach combines technology (an EHR that continuously validates documentation against PDPM coding requirements) with training, so that every clinician understands how their documentation affects classification, and with governance, so that coding decisions are subject to ongoing quality review. When all three elements are in place, the result is not just higher reimbursement. It is more accurate reimbursement: the kind that withstands audit scrutiny and reflects the genuine clinical complexity of the population you serve.

For skilled nursing facilities operating in an environment of rising costs and constrained margins, the difference between under-coded and accurately coded PDPM classification can represent hundreds of thousands of dollars in annual revenue. That revenue was already earned through the care your clinical team delivers every day. The only question is whether your documentation and coding practices are capturing it.