Dissatisfaction with therapy-driven reimbursement has been evident from many quarters of government for some time. In addition to multiple efforts by CMS to rein in therapy-driven reimbursement, reports from the Office of Inspector General in 2010, 2012 and 2015 identified therapy billing as a major driver of increased SNF reimbursement. A September 2015 report bore the title "The Medicare Payment System for Skilled Nursing Facilities Needs to Be Reevaluated." Multiple SNF providers have settled False Claims Act cases relating to the provision of therapy in SNFs. In March, MedPAC's 2017 Report to Congress contained a chapter titled "Post-acute care: Congress must act to implement recommended changes to PAC Payments" that emphasized the need for a payment methodology focused on resident characteristics. In introducing its proposed changes in the ANPRM, CMS echoed the concerns expressed in these examples, noting that "to better ensure that resident care decisions appropriately reflect each resident's actual care needs, we believe it is important to remove, to the extent possible, service-based metrics from the SNF PPS and derive payments from objective resident characteristics."
Changes to the Federal Base Rate Components
To accomplish this goal, CMS proposes several changes to the PPS. Throughout the ANPRM, it cites to an April 2017 report from an outside consultant that reflects the input of several Technical Expert Panels convened in 2015 and 2016. First, CMS proposes eliminating the therapy non-case-mix component of the federal base rate. The existing therapy case mix component would be bifurcated into a physical/occupational therapy component (PT/OT) and a speech language pathology (SLP) component. The nursing case mix component would be bifurcated into nursing and non-therapy ancillary services (NTA) components. The non-case-mix component would remain.
PT/OT Case Mix
Under the proposed approach, the PT/OT component would generate 30 case mix groups. Residents would no longer be assigned to a case mix group based on the amount of therapy provided. Instead, the assignment to a group would be based on patient characteristics that CMS has identified as predictive of therapy costs: (i) clinical reasons for the SNF stay; (ii) functional status; and (iii) cognitive impairment. CMS identifies ten clinical categories that CMS believes encompass the types of residents who use SNF services and further consolidates these into five categories of "clinical reasons" for a SNF stay that it believes predict PT/OT costs. These are Major Joint Replacement or Spinal Surgery, Other Orthopedic, Non-Orthopedic Surgery, Acute Neurologic and Medical Management. CMS proposes that data on the clinical reason for the SNF stay be drawn from item I8000 on the MDS 3.0.
Functional status would be evaluated based on the scoring of activities of daily living (ADLs) on the MDS, with adjustments to the scoring methodology to remove "bed mobility" as a scored category due to concern that facilities could drive reimbursement through this category (leaving transfers, eating and toileting as scored categories) and to use only items coded as "self-performed" to generate the RCS-I score. CMS notes its belief in this section that "self-performance items better represent the actual needs of the resident, while the support items represent facility resource decisions."
In evaluating cognitive status, CMS proposes a new scoring system, the Cognitive Function Scale (CFS), to replace the Brief Interview for Mental Status (BIMS) and Staff Assessment for Mental Status (SAMS) scores currently used.
Residents would be assigned to one of 30 PT/OT case mix groups based on their scores in each of the three categories.
SLP Case Mix
Like the PT/OT component, CMS would use scores on three predictors to generate 19 SLP case mix groups. For SLP, predictors would include: (i) clinical reasons for the SNF stay; (ii) presence of a swallowing disorder or mechanically-altered diet; and (iii) the presence of an SLP-related comorbidity. Residents with an Acute Neurologic diagnosis (based on the categories described above) would make up one clinical group and all other residents would be included in the other group. Residents would be categorized as having a swallowing disorder, as requiring a mechanically altered diet, or both. CMS also identifies several types of SLP-related comorbidities that would be scored.
Nursing Case Mix
CMS proposes to use the same 43 nursing case mix groups currently in use under the RUG-IV methodology, but it proposes to weight these groups so that classification into a nursing category is always part of the payment methodology for a resident. Currently, a resident is categorized into either a therapy or non-therapy RUG based on reimbursement rate. The case mix index for each nursing group would be calculated based on the average wage-weighted staff time per day for the group relative to the population average.
Non-Therapy Ancillary Case Mix
To remedy concerns that the current PPS does not adequately account for differences in non-therapy ancillary care costs, CMS proposes adding a separate component to be scored based on the presence of certain comorbidities and the need for extensive services. Comorbidities and services that would factor into the NTA score would be scored based on a weighted count. Examples of included factors for the NTA score include parenteral/IV feeding (high or low intensity), ventilator/respirator use, diabetes, IV medication, chemotherapy, transfusion and radiation. Notably, use of oxygen and BiPAP/CPAP were excluded from the list due to concerns that facilities could drive reimbursement through overuse. HIV/AIDS status would also factor into this score, but would be reported on the SNF claim rather than through the MDS due to state law conflicts.
Other Proposed Changes
In addition to changes to the federal base rate components, CMS proposes to add a variable per diem adjustment, to change the SNF assessment schedule, to address interrupted stays and to address the existing administrative presumption regarding qualification for a SNF level of care.
Per Diem Adjustment
CMS notes that resource use for PT/OT and NTA components tends to decline over the course of a SNF stay and remain constant for nursing and SLP components. Therefore, CMS proposes per diem adjustment factors based on payment day for the PT/OT and NTA components. The PT/OT case mix score would be adjusted down by one percent every three days beginning on day 15 of the resident's stay. By contrast, NTA score would be adjusted upwards by a factor of three for days one through three of a SNF stay and reduced to baseline for the remainder of the stay.
CMS proposes replacing the current schedule of multiple assessments throughout a resident's SNF stay with a single five-day MDS to set the RCS-I score and a PPS Discharge Assessment that would, among other things, track therapy minutes provided. Significant Change in Status Assessments would still be available for major status changes.
CMS also proposes to retain limitations on the amount of therapy that may be provided in a group setting to 25 percent of total minutes by discipline and to add a similar limitation to concurrent therapy.
Given the per diem adjustment proposed, CMS proposes that SNF stays interrupted by a hospital stay of more than three days would constitute a new admission for payment purposes, as would a transfer from one SNF to another. Readmissions to the same SNF after less than three days would not constitute a new admission.
Under current rules, certain administrative presumptions apply regarding qualification for the SNF level of care if a resident falls into a particular RUG category. Under the new proposal, the administrative presumption would apply for the same nursing categories to which it currently applies, as well as those PT/OT and NTA categories that reflect either high-intensity functional scores or high comorbidity scores.
Baker Donelson Comments
Comments on the ANPRM are due to CMS by June 26, 2017. SNF providers would be well advised to review the provisions of the rule in detail and to identify concerns early in the process.