Another Game Changing Proposal: The Patient Driven Payment Model

(August 2018 UPDATE note: CMS released the new Final Rule on July 31, 2018, which makes the info below out of date. Please click here to refer to our newer blog post and document dated August 3rd for the most up-to-date version of our PDPM interpretation.)

What a difference a year makes. One year and one day ago we posted about the "Game Changing" Proposed Final Rule, and here we are with a new one! Mark finished that update with, "Isn't Medicare Fun?" And it still is! To keep it fun, if you are awake all the way to the end, we’ll throw in some photos from last weekend when Mark digested all 266-pages of the new update and penned this summary. (If you are attending this week’s NARA Conference, the captions could prove helpful if your flight home is canceled - it’s a beautiful place about an hour from D.C. that would be fun to discover.) Without further ado, may we introduce you to the industry’s latest game changer, the Payment Driven Payment Model…

Initial Review of the Proposed Patient Driven Payment Model

Prepared by:  Mark McDavid, OTR, RAC-CT
Model’s Release Date:  April 27, 2018  •  Initial Review Release Date:  April 30, 2018

(Click Here for a PDF Version of this Review.)

Payment Model Background and Purpose of this Initial Review

As you may be aware, the Center for Medicare and Medicaid Services (CMS) published the 266-page Prospective Payment System and Consolidated Billing for Skilled Nursing Facilities (SNF) Proposed Rule for FY 2019 on Friday, April 27th. Included in the Rule is the annual payment update for SNF PPS rates, updates to the SNF Value-Based Purchasing Program and the SNF Quality Reporting Program, and the introduction of the Patient Driven Payment Model (PDPM). This document (the Proposed Rule) will be referred to as CMS-1696-P. You can find the full version of the Proposed Rule at:  Full Version.

In 2017, CMS published the Advanced Notice of Proposed Rule Making (ANPRM) which introduced the Resident Classification System – Version I (RCS-I) to the industry. The RCS‐I was developed as a result of the combined efforts of Acumen, a consulting group hired by CMS, and an interdisciplinary technical expert panel. Following publication, CMS began accepting comments. RCS-I would fundamentally change the way long-term care providers are reimbursed for the Part A post-acute care skilled services. For that reason, many different entities had varying opinions on RCS-I. As recently as the February 2018 SNF Open Door Forum (ODF), CMS was continuing to solicit comments on RCS-I even though the comment period was closed. This signaled to many of us that CMS was not happy with RCS-I in its proposed published form. In that ODF, John Kane, SNF Team Lead at CMS, said that there was no timeline for RCS-I. This, it turns out, foreshadowed RCS-I’s fate, which is that it has been scrapped (at least in part) for a new model by a different name.

This new proposed model, PDPM, was introduced Friday afternoon as part of this year’s SNF PPS Proposed Rule. This Initial Review was created after spending the last several days reading through and studying this release. The purpose of this Initial Review document is to discuss the important components of the new PDPM in order to educate readers on how SNFs will be reimbursed for Part A service if/when this new model goes from proposed to final rule.

The New Structure

Over the years, Medicare, MedPac, Congress, and various other stakeholders have been pushing to move the payment system away from counting therapy minutes to instead reimbursing SNFs based on patient characteristics and/or patient outcomes. The new PDPM will pay for SNF PPS-care based on patient characteristics and not a volume of services provided. The current RUGs-IV system is a case-mix index-maximizing system. In this system, as patients qualify for various RUGs (of the 66 RUGs available), then the system automatically assigns the highest paying RUG for each patient’s reimbursement. This is almost always a Rehab RUG (over 90 percent of SNF PPS days were reimbursed via one of the Rehab RUGs per CMS-1696-P). In the new PDPM, case-mix indexing will still play a role, but there are several different components that will contribute to the reimbursement that will ultimately equal the patient’s daily rate, or per diem. They are:  PT component, OT component, SLP component, Nursing component, Non-Therapy Ancillary component, and the Non Case-Mix component. The first five of which will be case-mix adjusted based on patient characteristics, and the last will be a flat rate.

Components 1 and 2:  Physical and Occupational Therapy Case-mix Classification

While Physical Therapy (PT) and Occupational Therapy (OT) will be separate components, their case-mix adjustment will be calculated together. There will be two patient characteristics used to determine the PT and OT case-mix classification. They are Clinical Category and Function Score. This process starts by using the clinical reason for the patient’s skilled stay to then place that patient into one of ten clinical categories. PDPM outlines 10 clinical categories based on ICD-10-CM diagnoses coded at I8000 on the MDS. Table 14 from CMS-1696-P outlines those 10 categories:

In the ANPRM document that outlined RCS-I last year, this was decreased to 5 categories for figuring the PT/OT component. However, based on comments to last year’s publication, CMS has now further pared down the PT and OT components to 4 categories in Table 15 of CMS-1696-P. These will be the 4 Clinical Categories driving the PT and OT case-mix classification:

This clinical category would be determined by the ICD-10-CM code reported on the first item of I8000 from the MDS 3.0. A review of ICD-10-CM mapping to clinical categories (https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SNFPPS/therapyresearch.html) will show that in some cases one ICD-10-CM will map to more than one clinical category. The multiple categories are due, in part, to the fact that resident needs will differ depending on whether or not the resident received surgical intervention while in the hospital immediately preceding this SNF stay.  Typically, patients who have had surgery in the immediately preceding hospital stay will require extensive post-surgical nursing or rehabilitation care in the SNF. If the patient did not receive a surgical procedure in the immediately preceding hospital stay, then the resulting clinical category will be a non-surgical one. Here is the example given in CMS-1696-P:

For example, certain wedge compression fractures that were treated with an invasive surgical procedure such as a fusion during the prior inpatient stay would be categorized as Major Joint Replacement or Spinal Surgery, but if these cases were not treated with a surgical procedure they would be categorized as Non-Surgical Orthopedic/Musculoskeletal. For residents who received a related surgical procedure during the prior inpatient stay, a provider would need to indicate the type of surgical procedure performed for the resident to be appropriately classified under PDPM.
— CMS-1696-P (Pages 81-82)

In order to capture the inpatient surgical procedure, CMS is requiring providers to use an ICD-10-PCS code corresponding to the surgical procedure in the second line on I8000.  If this proposal is finalized, then CMS will provide an ICD-10-PCS map of procedures to PDPM clinical categories.

The next step in determining the PT and OT component case-mix score is to determine the patient’s Function Score. Currently in the RUGs system (and previously proposed in RCS-I), CMS uses Section G: Functional Status (or the ADL score section of the MDS) to determine the patient’s functional score. In the new PDPM, CMS is proposing to use Section GG: Functional Abilities and Goals data to determine the Function Score. Thus, the Functional Score will be determined by four late loss ADLs and two early loss ADLs. Specifically, that includes two bed mobility items, three transfer items, one eating item, one toileting item, one oral hygiene item, and two walking items.  These were chosen as they are highly predictive of PT and OT costs per day. Tables 16, 17, and 18 of CMS-1696-P graphically show how these items will help produce the Function Score:

The scores for each of the Section GG items above will be added together to get a Total Function Score. This score will be used in Table 21 to further place the patient into the appropriate Case-mix Classification Group.

As you may recall, the RCS-I proposal included cognitive status as a determinant for the PT/OT case-mix classification. Due to comments in response to ANPRM, CMS has since removed cognitive status as a determinant of the PT and OT case-mix classification. CMS uses a Classification and Regression Trees (CART) algorithm to help determine the resident groups and cognition was included as an independent variable in the CART analysis. For more information on the CART analysis refer to page 93 of the CMS-1696-P document.

Taking 1) Primary clinical reason for SNF stay and 2) Function Score into account, we can now classify a patient into a PT and OT Case-Mix Classification Group. Table 21 graphically shows the breakdown of each PT and OT Case-mix Classification Group:

Component 3:  SLP Case-mix Classification

Speech Language Pathology case-mix classification will follow the same methodology as the PT and OT classification.  Initially, it will be determined if the primary clinical reason for the SNF stay is either “Acute Neurologic” or “Non-Neurologic”.  A quick review of Table 14 above shows that “Acute Neurologic” has its own category and “Non-Neurologic” would be comprised of the 9 remaining categories.

The second characteristic to be identified for the SLP Case-mix Classification is the presence of a swallowing disorder or mechanically altered diet. It is noted that there is increased provider cost for residents with either a mechanically altered diet or a swallowing disorder. And a further increase in cost if the resident has both. Therefore, “both”, “either”, or “neither” are used as it relates to swallowing disorder and mechanically altered diet when determining the SLP Case-mix Classification.  These would be identified by responses to K0100Z (swallowing disorder) and K0510C2 (mechanically-altered diet) on the MDS.

The third characteristic to be identified for the SLP Case-mix Classification is cognitive status or whether or not the resident had a SLP-related comorbidity present. Table 22 shows the SLP-related Comorbidities:

As with the RCS-I ANPRM proposal, CMS is proposing to use the Cognitive Functional Scale which is a combination of the Brief Interview for Mental Status (BIMS) and Cognitive Performance Scale (CPS) to identify cognitive status for the SLP Case-mix Classification. Due to comments to the ANPRM proposal, CMS has moved a score of “0” on the CPS to equal “Cognitively Intact” vs. “Mildly Impaired”. Table 20 graphically shows how cognition level is chosen using these two tools:

Finally, CMS was able to decrease the number of SLP Case-mix Classification Groups from 18 (in the ANPRM proposal) to 12 for PDPM purposes. Table 23 shows the breakdown of how SLP Case-mix Classification Groups will be assigned:

Component 4:  Nursing Case-Mix Classification

CMS is proposing to identify the Nursing Case-Mix Classification based on a modified traditional RUG-IV methodology and decreasing the possible RUGs from 43 to 25. In the traditional RUG-IV nursing RUG methodology, the ADL score was derived from Section G of the MDS. Under PDPM, CMS is proposing to use Section GG to determine a Function Score as outlined below in tables 24 and 25:

Decreasing the number of RUG options from 43 to 25 was accomplished by collapsing case-mix groups that had contiguous ADL scores when those RUGs were defined by similar clinical traits. Table 26 outlines which RUGs were collapsed and identifies the Nursing case-mix index score:

HIV/AIDS Add-on

CMS explained that there was a significant increase in nursing cost to provide care for residents with HIV/AIDS. To compensate facilities for caring for these residents, CMS proposes an 18% increase in the nursing component. This increase would be applied based on the presence of ICD-10-CM code B20 on the SNF claim.

Component 5:  Non-Therapy Ancillary

Non-Therapy Ancillary (NTA) scoring is fairly straight forward. Using Table 27 as a reference, if the resident has the condition/extensive service on the left column from the noted source location, then sum the points identified on the right together to get the total NTA score:

Based on the NTA scores summed from the conditions/extensive services on Table 27, Table 28 reflects the NTA Case-Mix Classification Groups:

Component 6:  Non Case-Mix Component

The final piece of the total per diem is the non case-mix component. This flat-rate component (that is not case-mix adjusted) is for covering room and board, capital expenses, and administrative overhead.

Combining all 6 components above (PT, OT, SLP, Nursing, NTA, Non Case-Mix) will result in the base daily rate for that patient. Multiplying each component’s base rate by its respective case-mix index will return the facility’s daily rate for that specific resident upon admission. As will be discussed below, the Variable Per Diem Adjustment Factor comes in play with PT, OT and the NTA components. These components will be decreased at various intervals during the resident’s stay.

Variable Per Diem Adjustment Factors

Discussed in both the ANPRM and CSM-1696-P documents, is Medicare’s discovery of decreasing costs of PT, OT, and NTA the longer the resident stays in the facility. This discussion can be found throughout the CMS-1696-P document. Due to the decreasing costs during a resident stay, CMS proposes the Variable Per Diem Adjustment Factor.

For PT and OT, the PDPM proposes that the facility be reimbursed the full base rate multiplied by the the case-mix index, for each resident for PPS days 1-20. Then, starting on day 21, a decreasing adjustment factor of 2% every 7 PPS days would be applied.   

  • Days 1-20: PT base rate x CMI x 1.00 (the adjustment factor) = PT portion of the per diem
  • Days 21-27: PT base rate x CMI x .98 (the initial adjustment factor) = PT portion of the per diem.

This decrease would continue for both PT and OT every 7 days as outlined in Table 30 until the patient is discharged. It is worth noting that if the resident stayed all 100 days, the PT and OT components would be reimbursed at 76% of their initial rate for days 98-100.

For the NTA, the proposed adjustment factor starts at a multiple of 3 and is reduced to a multiple of 1 on day 4 as outlined below:

Resident Assessment Instrument:  MDS, Version 3 Proposed Changes

As CMS is no longer using therapy minutes to determine reimbursement, any changes in a resident’s therapy minutes will no longer impact reimbursement, therefore, CMS is proposing a new MDS Assessment schedule. The new PPS MDS schedule is represented in Table 33:

Interim Payment Assessment (IPA)

An IPA would be required to be completed where the following two criteria are met:

(1) There is a change in the resident’s classification in at least one of the first tier classification criteria for any of the components under the proposed PDPM (which are those clinical or nursing payment criteria identified in the first column in Tables 21, 23, 26, and 27) such that the resident would be classified into a classification group for that component that differs from that provided by the 5-day scheduled PPS assessment, and the change in classification group results in a change in payment either in one particular payment component or in the overall payment for the resident; and

(2) The change(s) are such that the resident would not be expected to return to his or her original clinical status within a 14-day period.

In addition, we propose that the Assessment Reference Date (ARD) for the IPA would be no later than 14 days after a change in a resident’s first tier classification criteria is identified. The IPA is meant to capture substantial changes to a resident’s clinical condition and not every day, frequent changes. We believe 14 days gives the facility an adequate amount of time to determine whether the changes identified are in fact routine or substantial.
— CMS-1696-P (page 140)

In last year’s ANPRM document, CMS proposed keeping the Significant Change assessment and is now proposing that the Significant Change assessment be replaced by the IPA.

CMS has also proposed a change in the rules governing when a PPS Discharge Assessment is to be completed and the Assessment’s Item Set. Currently, the PPS Discharge Assessment is completed when a patient is discharged from Part A and continues to stay in the facility. CMS is proposing the completion of the PPS Discharge Assessment for all residents who are discharged from PPS; whether they are actually discharged from the facility or stay long term. Additionally, CMS is proposing to change the Discharge Assessment Item Set to include a modified Section O.  

In CMS-1696-P, CMS discusses the comments received in response to the ANPRM publication where stakeholders are concerned about the lack of oversight and monitoring of therapy provisions throughout the stay of the resident.  Therefore, CMS now proposes to monitor therapy provisions throughout the stay by adding Table 35 to Section O on the SNF PPS Discharge Assessment:

It appears that the added “Total” items (total individual minutes, total concurrent minutes, total group minutes, total days) may be referring to totals throughout the resident’s stay and not just in the 7-day look-back period. As of this writing, this has not been confirmed and the CMS-1696-P document is not explicitly clear on this issue.

Grace Days

Note: In the current MDS 3.0, RUGs-IV Model the 5-day assessment window is days 1-5 with allowable grace days 6-8. CMS is proposing the removal of the label “Grace Days” from the 5-day assessment window. The scheduled assessment window for the 5-day assessment will now be identified as days 1-8 of the PPS stay.

Therapy Provision Policy Changes

Even though we will no longer be using therapy days, modes, and minutes to determine payment, CMS will continue to track these items and will be placing additional constraints on therapy providers. It is expected that the majority of therapy provided will be in the individual mode. To that end, CMS is expanding the group limitation to include concurrent therapy as well. That is, CMS is proposing limiting group and concurrent treatment provided to a resident to 25% of the minutes provided per discipline. This, again, was in response to stakeholder comments where there was concern that a large portion of therapy would now be provided in a group or concurrent setting. Additionally, CMS is going to allow the entire provision of minutes provided in group and concurrent to be recorded on the MDS, which differs from the current RUGs-IV system where the Reimbursable Therapy Minutes are calculated by the group (one fourth) and concurrent (one half) reduction.

Should your therapy department provide more than the allowed 25% of group and concurrent per discipline, the provider would receive a non-fatal warning edit on the validation report after submitting an MDS to the QIES ASAP system. This would serve to alert the provider that the modes utilized in the provision of therapy were such that they exceeded the 25% limitation. Note, this would not be a fatal error which indicates that one or more submitted items failed to pass the requirements identified in the MDS. CMS indicates that they will monitor group and concurrent utilization and consider making future proposals to address abuses of this policy or flag providers for additional review.

Interrupted Stay Policy

The interrupted stay policy outlined in last year’s ANPRM lives on in its proposed form. The interrupted stay policy is proposed to relieve any incentive for providers to discharge a resident with the intent of bringing them back into the facility in a few days in order to reset the Variable Per Diem Adjustment Factors. In the interrupted stay policy, a resident’s PPS calendar will resume with the next PPS day if the resident returns to the facility within 3 midnights (the midnight of the day of discharge counts as one of these three). In cases where the resident returns within this timeframe, there would be no new 5-day Assessment, nor would the Variable Per Diem Adjustment Factor be reset. The resident would return and the payment schedule would continue on the next PPS day continuing with the Variable Per Diem Adjustment Factors in place. This only applies if the resident is out of the facility for less than 3 days. If the resident returns on day 4 or later or is sent to a different facility, then the Variable Per Diem Adjustment Factors are to be reset to day 1 and a 5-day assessment would be required.

A PDPM Example

In an effort to help make the proposed PDPM clearer, let’s take a look at an example of how the per diem will be calculated under the PDPM: Mr. B is a hip replacement patient and has a PT and OT Case-Mix Group of TB, SLP Case-Mix Group of SA, Nursing PDPM Case-Mix Group of CDE2, NTA of NE, and a flat non case-mix rate. The following calculations determine the facility’s per diem covering Days 1-3 for Mr. B’s care.

Due to the Variable Per Diem Adjustment Factors, the NTA changes on day 4 and is constant throughout the rest of the stay. The PT/OT Component decreases by 2% every seven days beginning on day 21. These adjusted rates for our example patient, Mr. B, are listed below:

For Days 4-20, the NTA is 2/3 less:

  • Urban - $565.35 per day
  • Rural - $584.25 per day

Starting on Day 21, the PT and OT Components decrease by 2% every 7 days. This results in the following per diem for days 21-27:

  • Urban - $561.54 per day
  • Rural - $579.94 per day

Comments

As you can see, the proposed PDPM is a vast change from the current RUGs-IV system.  Over the coming weeks, Mark will be working with NARA and AANAC to ensure that your concerns about the proposed payment model are made known to CMS via the comment process.  Should you or your facility like to comment directly to CMS, they are soliciting comments in the following ways:

1.  Electronically. You may submit electronic comments on this regulation to http://www.regulations.gov.

Follow the "Submit a comment" instructions.

2.  By regular mail. You may mail written comments to the following address ONLY:

Centers for Medicare & Medicaid Services,
Department of Health and Human Services, Attention: CMS-1696-P,
P.O. Box 8016,
Baltimore, MD 21244-8016.
Please allow sufficient time for mailed comments to be received before the close of the comment period.

3. By express or overnight mail. You may send written comments to the following address ONLY:

Centers for Medicare & Medicaid Services,
Department of Health and Human Services,
Attention: CMS-1696-P,
Mail Stop C4-26-05,
7500 Security Boulevard,
Baltimore, MD 21244-185

Closing Note

As you may be thinking, this new model will have large implications for all types of long-term care providers. The review you have just read comprised our initial summary of PDPM, but as we spend more time with the model other issues and nuances will come to light over the coming months. PDPM is something that we, as an industry, will discuss and address in order to successfully navigate the changing regulations. 

We will be doing just that in various settings over the next year and a half. If you are not already a regular client, please visit our website (seagroverehab.com) and sign up to be informed about any updates. We will most likely host a webinar on this topic, and if you attend any therapy-related conferences, keep an eye out for Mark’s sessions. He is a scheduled speaker at American Health Tech’s customer symposium this May and may be speaking at other state and national association meetings near you (i.e. he has spoken at AANAC’s annual conference the last several years).

In the meantime, should you or someone in your facility have questions about the new model, please feel free to reach out to us. We welcome the opportunity to create a partnership between our teams.

Mark McDavid, OTR, RAC-CT

850-532-1334

mark@seagroverehab.com

(To view this review as a PDF, click here.)

About Seagrove Rehab Partners

Seagrove Consulting Group, LLC (dba Seagrove Rehab Partners) is a national healthcare consulting company headquartered in Santa Rosa Beach, Florida. Our mission is to serve the therapy industry by providing a compassionate, people-centered approach to compliance and management support, including a proprietary solution for in-house therapy programs. We believe that a thoughtful strategy not only helps our partners to be more compliant and profitable, but also puts the patient first. Mark McDavid, president, and our 10+ company associates are available to work with skilled nursing facilities, rehabilitation companies, and other healthcare providers at the client site, by phone, by email, by video conference call and through workshops and seminars.


The fun stuff... if life takes you to Virginia, we highly recommend the Inn at Vaucluse Spring. The new owners are wonderful and the setting is serene. Perfect for lots of Medicare Reading! :)