Reed Smith Client Alerts

  1. INTRODUCTION

    On May 12, 1998, the Health Care Financing Administration ("HCFA") issued an interim final rule with comment period implementing provisions of the Balanced Budget Act of 1997 ("BBA") that established a prospective payment system ("PPS") and consolidated billing requirements for skilled nursing facilities ("SNFs"). 63 Fed. Reg. 26,252.

    Specifically, this rule implements section 4432 of the BBA, which provides that, for cost reporting periods beginning on or after July 1, 1998, SNFs will no longer be reimbursed retrospectively for the reasonable cost of services provided, but rather will be paid under a PPS consisting of a per diem rate based on the intensity of resources required by each resident. In addition, each SNF will be required to "consolidate" its billing, by submitting a single bill to Medicare for all items and services provided to its residents, with certain exceptions, irrespective of whether the item or service is payable under Part A or Part B, and regardless of whether the service was provided by the SNF or by an outside supplier or provider. These two components of the interim rule, the PPS and consolidated billing, are discussed in detail below.

    We note that this interim rule is quite lengthy, and we therefore do not attempt to discuss every aspect of the rule. In particular, the rule includes an extensive discussion of HCFA’s design and methodology for developing the case-mix adjusted federal rates and various indices that will be used to adjust and update the rates. The analysis of these data and indices is beyond the scope of this memorandum, although we anticipate there will be considerable industry commentary on the data and methodologies used to implement the new PPS system. Instead, in this memorandum, we seek to underscore what we believe to be the most important components of the PPS and consolidated billing for SNFs and entities providing services to SNF residents, and thereby alert our clients to issues that could affect their business. Please let us know if you need additional information regarding any of these issues. HCFA will accept comments on the interim rule received by July 13, 1998.

  2. THE PPS FOR SNFs
    1. Background/Summary

      Presently, SNFs are reimbursed by Medicare retrospectively for the reasonable cost of covered services. This type of reasonable cost reimbursement applies to routine, ancillary, and capital-related costs. Section 4432(a) of the BBA amended the Social Security Act (the "Act") to phase-out this reasonable cost reimbursement payment system and institute a payment system whereby SNFs are reimbursed for all covered services at a prospectively-determined per diem rate. This PPS is statutorily mandated to be phased-in over three cost reporting periods, beginning with cost reporting periods beginning on or after July 1, 1998. Because of this statutory mandate, HCFA has indicated it is committed to timely implementation, even if all aspects of the PPS are not fully operational.

      Prospective payment rates include all costs of furnishing covered skilled nursing services (i.e., routine, ancillary, and capital-related costs), except that costs associated with operating approved educational activities are excluded and are separately reimbursable. Services covered by the PPS rate include post-hospital SNF services provided under Part A, and all items and services (except for certain statutorily excluded services) provided under Part B that are furnished during a Part A covered stay.

      The SNF PPS is a clinically-driven reimbursement system, based upon rates determined, essentially, from historical costs. Under the PPS, payment will be made according to the quantity and quality of resources required by a particular resident. Accordingly, all residents are categorized into one of 44 Resource Utilization Groups ("RUGs") as established in Version III of RUG ("RUG-III"). Significantly, RUG assignment is based on resource consumption, i.e., the cost of items and services needed to care for the particular resident. Therefore, reimbursement under the PPS for SNFs is not a direct function of the patient’s diagnosis as is the case for the PPS for hospital services, which is based on Diagnosis Related Groups ("DRGs"). Rather, as explained in greater detail below, the PPS for SNFs utilizes RUG-III groupings, which are intended to reflect the intensity of care and services required to meet the resident’s care needs based upon a comprehensive assessment.

      The PPS will be phased-in incrementally over a period of three cost reporting periods (not years), during which most SNFs will receive a payment comprised of both the federal rate and a facility-specific rate. Over the transition period, this rate will shift towards a greater proportion of the federal rate, and a lesser proportion of the facility-specific rate, until in the fourth cost reporting period, SNFs will be reimbursed only at the federal rate. SNFs that received their first Medicare payment on or after October 1, 1995 (under present or prior ownership) will initially receive payment at the fully federal rate, i.e., these facilities will not participate in the transition period. A flow chart which provides a general overview of the PPS, from the resident’s day of admission to payment from the intermediary, is attached as Exhibit A.

      In total, the advent of the PPS for SNFs signals a dramatic shift in the health care industry, as SNFs become the consumers/purchasers of items and services and must seek to streamline the provision of quality care and at the same time implement cost-saving efficiencies in order to fit within the per diem rates. Moreover, the PPS, and its reliance on clinical documentation as the basis for reimbursement through utilization of the RUG-III classification system, shifts the responsibility for procuring reimbursement from billing personnel to clinicians who complete the Minimum Data Set ("MDS") resident assessment tool. In this way, the PPS dramatically expands the role of clinicians and heightens their importance as participants in the Medicare reimbursement system.

    2. The Federal Payment Rates

      The federal payment rates are based on mean SNF costs in a base year, subject to certain adjustments, and updated for inflation to the first effective period of the PPS. The rates are developed using allowable costs from hospital-based ("HB") and freestanding ("FS") SNFs from cost reports for cost reporting periods beginning in federal fiscal year ("FY") 1995 (i.e., beginning on or after October 1, 1994 through September 30, 1995). The federal rates also incorporate an estimate of the amounts payable under Part B. Rates are updated using a SNF market basket index ("MBI") and are standardized to allow for facility differences in case-mix and wages. Cost data for payments made to "new providers" and other qualifying providers pursuant to exceptions to the routine cost limits ("RCLs") are not included in the database used to compute the federal rate. As required by the BBA, each rate is set at a level equal to a weighted mean of the FS SNF costs plus 50 percent of the difference between the FS SNF mean and a weighted mean of all SNF costs (FS and HB) combined. Rates are computed and applied separately for facilities in urban and rural areas.

      1. Data Sources Used To Develop The Federal Payment Rates

        While the federal rates generally are derived from FY 1995 SNF cost reports, the data are subject to numerous complex adjustments and indices intended to account for case-mix variation, geographic variation, and inflation. Specifically, the data utilized in developing the federal rates was gathered from five different sources: (1) FY 1995 cost reports; (2) Part B claims data; (3) a wage index; (4) case-mix indices; and (5) a market basket inflation index. Each is discussed generally below.

        Cost report data was gathered from both HB and FS SNFs for FY 1995. Only cost reports for periods of at least 10 months, but not more than 13 months, were included in the database because, according to HCFA, the shorter and longer cost reporting periods may not reflect a "normal" cost reporting period. 63 Fed. Reg. at 26,256. In addition, providers that were exempted from RCLs for FY 1995 were excluded. Allowable costs related to RCL exceptions and costs related to approved education activities also were excluded from the calculation.

        Part B charges were estimated by matching 100 percent of the Medicare Part B SNF claims associated with each Part A resident’s covered SNF stay to the SNF cost reports. The allowable charges were then incorporated at the facility level in the appropriate cost center (e.g., laboratory, medical supplies) with the cost report data. While it is somewhat unclear from the interim final rule whether HCFA also matched Part B charges for Part A SNF residents submitted by outside providers and suppliers, HCFA has stated that the Part B match includes such claims, and is intended to be complete.

        Data necessary to compute a wage index for SNFs was not available, and consequently the hospital wage index was utilized. The industry previously has challenged the use of the hospital wage index in the context of SNF reimbursement, such as in calculating the salary equivalency guidelines for occupational and speech therapy services. HCFA now is required by Congress to develop a SNF wage index, and once that index is developed, HCFA could (but is not required) use the SNF wage index for future rate adjustments. However, industry sources speculate that even when a SNF wage index is developed, it will not necessarily increase the total amount of funding to the system, but rather shift funds between facilities in different geographic areas.

        The interim final rule uses two different data sources to standardize the federal cost report data for variation in case-mix. (The regulation itself contains a vague statement that "cost report data are standardized for facility variation in case-mix using case-mix indices and other data that indicate facility case-mix." 42 C.F.R. § 413.337(b)(4).) First, as noted, the PPS uses weights derived from the RUG-III resident classification system. These are based on nursing and therapy time related to residents in each of the RUG groupings. In other words, data from the RUG-III system is used to standardize the federal rates and for the case-mix adjustments to the rates paid for particular residents based on their different service needs. (The RUG-III system is discussed in greater detail in Section II(C), below.)

        Second, HCFA utilizes a standardization device to remove the effects of geographic variation in case-mix called the Medicare Provider Analysis and Review ("MEDPAR") analog. According to HCFA, this database contains clinical information from FY 1995 and was utilized because a national source of MDS-derived case-mix data for this period does not exist. 63 Fed. Reg. at 26,257. Accordingly, HCFA constructed a crosswalk between the RUG-III categories and data in the MEDPAR file. In this way, HCFA states that it was able to provide national case-mix data on SNFs in its database. Id.

        Finally, HCFA uses the SNF MBI to develop the initial federal payment rates (i.e., update the FY 1995 data to 1998 cost reporting periods), and also to update the federal rates on an annual basis beginning in FY 2000. A discussion of the SNF MBI is set forth in Section II(F), below.

      2. Calculation Of The Federal Payment Rates

        The methodology used to compute the per diem federal rates is a complex, multistep process utilizing each of the five data sources discussed above. This calculation results in the "Unadjusted Federal Per Diem Rate," consisting of four cost components: (1) nursing -- case-mix; (2) therapy case-mix; (3) therapy -- non-case-mix; and (4) non-case-mix. These cost components are computed separately for urban and rural SNFs. 63 Fed. Reg. at 26,260. These cost components are, in turn, adjusted by the RUG-III case-mix grouping payment index to determine the "case-mix federal rate," i.e., the final federal rate.

      3. Examples Of Federal Payment Rates

        The federal rates effective for the first 15-month federal year (i.e., July 1, 1998 through September 30, 1999) are published in the interim final rule discussed herein. As discussed more fully below, given that the federal payment rates are intended to reflect differences in the amount and type of resources required to care for each payment, they vary considerably. For example, with respect to federal payment rates for urban SNFs, the rate for the highest RUG category (RUC -- Rehabilitation major patient type, Ultra High category) is $384.21. The lowest payment rate (same as the default rate), which is in the Reduced Physical Function category, is $117.15. Recall, however, that many patients assigned to this lowest rate will not qualify for Medicare. With respect to federal payment rates for rural SNFs, the highest rate is $408.19 and the lowest is $116.85.

      4. Updating The Federal Rate

      As noted above, for the initial period of the PPS beginning on July 1, 1998, and ending on September 30, 1999, the payment rates are set forth in the regulation published on May 12, 1998, and discussed herein. For each subsequent fiscal year, HCFA will publish the updated federal rates in the Federal Register before August 1 of the year preceding the affected federal fiscal year.

      Regarding FYs 2000 through 2002, the federal payment rates will be increased by a factor equal to the SNF MBI change minus one percentage point. All subsequent years will be updated by the applicable SNF MBI.

      Finally, HCFA notes that it has received "suggestions" from other oversight entities -- presumably the Office of Inspector General ("OIG") -- that the federal rates should be adjusted downwards to "reflect the presence of inappropriate care or payments in the 1995 cost data used to establish" both the federal and facility-specific rates. 63 Fed. Reg. at 26,267. Furthermore, HCFA states that concerns have been raised regarding whether these data are inflated, reflecting medically-unnecessary care and/or payments related to therapies and other ancillary services. Id. Lastly, HCFA states that studies conducted by the OIG and HCFA program integrity activities have found that incorrect payments have been made to SNFs in the past, but that the magnitude of the incorrect payments is not known. Accordingly, the OIG, in conjunction with HCFA, plans to investigate this matter further and, if excessive inappropriate costs are revealed, HCFA plans to address that issue in another rule, or perhaps seek legislation to adjust future payment rates downward. Id. Numerous providers are expected to comment on this issue and we anticipate that the industry as a whole will strongly oppose any downward adjustment to the payment rates.

    3. Resource Utilization Group, Version III ("RUG-III")
      1. Introduction

        The federal rate is responsive to the case-mix (the relative resource utilization of residents) of each facility by determining payments according to 44 RUGs. As noted, RUG-III is a classification system that categorizes individual patients according to the amount and type of resources they utilize. The RUG-III classification system adjusts for case-mix in that each RUG is tied to a payment index used to establish variable payment levels. The index score keyed to each RUG is intended to represent the amount of nursing time and rehabilitation treatment time associated with caring for the patients who are assigned to that particular group.

        Under the RUG-III system, SNFs are required to utilize the resident assessment data from the MDS Version 2.0 in order to assign (via use of the "Grouper" software, discussed below) each resident to one of 44 RUGs. The MDS contains extensive clinical information compiled by the SNF and contains all data necessary to classify a resident into a specific RUG. First, each resident is assigned one of the seven major patient types based on information contained in the MDS. Second, within each major patient type, each resident is assigned to one of the resulting 44 RUGs. 63 Fed. Reg. at 26,257.

        MDS information used for RUG classification include diagnoses, ability to perform activities of daily living ("ADLs"), and treatments received. Exhibit B schematically depicts how information contained on the MDS is used to assign each resident to one of 44 RUG categories. Also, the RUG-III classification system establishes automatic Medicare eligibility for residents in the top four major patient types, as discussed further below.

      2. Seven Major Patient Types

        As noted above, under the RUG-III classification system, each resident is initially assigned to one of seven hierarchical major patient types, in order from most intense to least intense: (1) Rehabilitation, (2) Extensive Services, (3) Special Care, (4) Clinically Complex, (5) Impaired Cognition, (6) Behavior Problems, and (7) Reduced Physical Function. These major patient groups are further differentiated into the 44 RUGs, to which all patients are assigned. The following is a brief summary of each of the seven major patient types.

        a. Rehabilitation. This is the most complex major patient type. It has five subcategorizes: (1) Ultra High; (2) Very High; (3) High; (4) Medium; and (5) Low. These five subcategories are further differentiated to yield 14 corresponding RUG categories.

        Assignment to the Rehabilitation category is based on the number of minutes of therapy provided, the number of disciplines providing therapy, and the number of days therapy is provided per week. 63 Fed. Reg. at 26,258. For example, a resident in the "Ultra High" category will receive at least 720 minutes of therapy a week, while a resident in the "Medium" category will receive a minimum of 150 minutes.

        The next subcategory in the Rehabilitation group is determined by the resident’s ADL score. ADL scores range from 4 to 18. A high score indicates that the patient needs significant assistance, training, and rehabilitation in order to complete the activities of daily living. While there are numerous ADLs reported on the MDS, only four are utilized for purposes of RUGs classification. These are: (1) bed mobility; (2) toilet use; (3) transfer from bed to chair; and (4) eating. In this category, the resident’s ADL score places them into one of the fourteen rehabilitation categories or RUGs. 63 Fed. Reg. at 26,262.

        b. Extensive Services. Residents qualify for this category if: (1) in the past 14 days they received intravenous medications, tracheotomy care, required a ventilator/respirator, required suctioning; or (2) in the past seven days, they received intravenous feeding. Patients assigned to this major patient type category will have an ADL score of at least seven.

        The patient is assigned to one of three RUGs within this category on the basis of five criteria. The first three criteria are presence of a clinical condition that qualifies the patient for classification to the (1) Special Care; (2) Clinically Complex; or (3) the Cognitively Impaired major patient type categories. The remaining criteria are (4) whether the patient is receiving intravenous feeding; or (5) whether the patient is receiving intravenous medication. In sum, the RUG subdivisions in this category are based on the technical clinical knowledge and skill required to care for a particular patient.

        c. Special Care. Patients who are assigned to this category have at least one of the following: multiple sclerosis; cerebral palsy; quadriplegia with an ADL score of 10 or more, or receive respiratory therapy seven days per week; have, and receive treatment for, pressure or stasis ulcers on two or more body sites; have a surgical wound(s) or open lesions; receive tube feedings with at least 26 percent of daily calorie requirements and at least 501 ml of fluid through the tube per day, and are aphasic; receive radiation therapy; or have a fever in combination with dehydration, pneumonia, vomiting, weight loss, or tube feedings.

        After residents are assigned to a special care major patient type category, they are assigned to one of the three corresponding RUGs based on their ADL score.

        d. Clinically Complex. Patients qualify for this category if they are comatose, have burns, septicemia, pneumonia, internal bleeding, dehydration, dialysis, hemiplegia in combination with an ADL score of 10 or more, receive chemotherapy, receive tube feedings that comprise at least 26 percent of daily calorie requirements and at least 501 ml of fluid through the tube per day, receive treatments for foot wounds, or transfusions. 63 Fed. Reg. at 26,264. Although it is unclear from the language of the rule, we assume that a resident need not present with each of these symptoms to be assigned to this major patient type; instead we believe one of these maladies would be sufficient.

        Also included in this category are diabetics who receive injections seven days per week and who have two or more physician order changes in the past 14 days, as well as patients who have received oxygen therapy in the past 14 days. In order to assure inclusion of patients with unstable conditions, HCFA also used a combination of physician visits and order changes as qualifying criteria for this category. This is a proxy measure for the amounts of skilled nursing observation, care planning, and monitoring usually required by this type of patient. The qualifying combinations of physician visit/order changes that must occur within the 14 day observation period to qualify for this category are: (i) one or more visits with at least four order changes, or (ii) two or more visits with two or more order changes. Patients are assigned to one of six RUGS with this major patient type based on whether they show signs of depression.

        e. Impaired Cognition. Patients in this category, as well as each of the following categories, frequently will not qualify for Medicare coverage. In other words, the Impaired Cognition major patient type category marks the beginning of the 18 RUGs that fall below the line of per se Medicare eligibility (discussed below).

        The patients in this category will have scores on the MDS 2.0 Cognition Performance Scale of 3, 4, or 5, and for two of the groups in this category will be receiving nursing rehabilitation services six days per week. Some patients with Alzheimer’s disease or other types of dementia who have been acutely ill will be classified in this category and still qualify for Medicare. Under the SNF coverage guidelines, these patients could qualify based on the need for skilled nursing rehabilitation. The Impaired Cognition major patient type is differentiated into four RUGs, based on the amount and type of skilled nursing services required by the patient.

        f. Behavior Only. These are patients who, in four of the last seven days, exhibited behaviors that include resisting care, being combative, being physically and/or verbally abusive, wandering, and who have hallucinations or delusions. Similar to the Impaired Cognition category, this category is differentiated or "end-split" into four different RUGs based on the skilled nursing services required.

        g. Physical Function Reduced. The patients in this category are those who do not have any of the conditions or characteristics identified above. However, some have been documented as receiving "skilled nursing" and have been covered by Medicare in the past. In the preamble to the regulations, HCFA indicates that with proper documentation and justification regarding the need for skilled care, Medicare may continue to cover SNF services. 63 Fed. Reg. at 26,264. However, SNFs may justifiably be skeptical of this assertion in light of the fact that this category is the lowest in the hierarchical RUG-III classification system, far below the per se Medicare eligibility line.

      3. Summary Of RUG Assignment Process

        As noted above, each of the seven major patient types is further differentiated to yield the 44 specific patient groups used for payment. For all major groups except Rehabilitation and Extensive Services, the first differentiation or subdivision (downward from the major patient types), is based on the patient’s ADL score.

        The next, and last subdivision (sometimes referred to as "end-split") is based on nursing rehabilitation services or signs of depression. 63 Fed. Reg. at 26,263. The initial subdivision for the Rehabilitation category is the number of minutes per week of rehabilitative therapy services, the number of rehabilitation disciplines providing services, and the number of days per week therapy is provided. The second subdivision uses the patient’s ADL score. The Extensive Services category uses the resident’s ADL score for threshold assignment to the category, but subdivisions within that category are based on the technical clinical knowledge and skill required.

      4. RUG-III "Grouper" Software

        All data necessary to assign a resident to one of the RUG-III categories is contained in the MDS. Software programs known as "groupers" utilize information from the MDS to assign patients to the appropriate RUG. HCFA’s grouper is available free of charge on its web site: www.hcfa.gov/medicare/nsqb/mds20.

        It is important to note that the grouper always assigns patients to the highest weighted group, rather than just the higher group in the hierarchy. This is significant only in the rare instance where a patient would qualify for a group that is higher in the RUG-III hierarchy, but has a lower payment index than a group that is lower in the hierarchy. However, in practical terms this means that SNFs using the grouper should be assured that the resident will always be assigned to the higher paying RUG. See 63 Fed. Reg. at 26,264.

      5. The MDS And The Importance Of Clinical Documentation

        The MDS is a form that contains numerous alphabetical sections, each of which contain several questions regarding the resident and his or her medical history and present conditions. Historically, the MDS has been part of each state’s standardized, comprehensive Resident Assessment Instrument ("RAI") for purposes of conducting the initial and periodic resident assessments of each resident’s functional capacity. See 42 C.F.R. § 483.20. Thus, SNFs are familiar with the MDS and the resident assessment process. The most recent electronic version of the MDS is 2.0. HCFA provides access to the MDS data entry software, called "Raven," on its web site free of charge.

        Under the PPS, SNFs will be required to complete two additional sections of the MDS, Sections T and U. Section T contains information on special treatments and therapies not reported elsewhere in the patient assessment. SNFs must complete this section for services furnished on or after July 1, 1998. Section U currently is an optional section used to collect information on medication. SNFs do not have to complete and transmit this information until October 1, 1999. 63 Fed. Reg. at 26,265.

        Given that SNF residents are assigned to RUG-III classifications based on information contained in the MDS, the information contained in the MDS effectively drives reimbursement. Moreover, the fact that clinicians normally complete the MDS places heightened importance on their role in SNF reimbursement. In sum, the PPS for SNFs is a self-reporting system and, therefore, the accuracy of documentation is extremely important. Great care must be taken to ensure that each patient’s conditions and treatments are fully documented, and that information in the patient chart matches information on the MDS. In addition, SNFs must ensure that the MDS for each resident is internally consistent. For example, it would be inappropriate for the MDS to reflect 500 minutes of speech language therapy per week when the MDS indicates, in another section, that the patient is in a coma.

        SNFs may also want to consider revising their corporate compliance programs and policies in order to reflect the increased significance placed on the accuracy and completeness of documentation. For example, SNFs should consider adopting a policy on timeliness, specificity, and accuracy of the MDS, as well as instituting a facility-wide training program on the PPS, with special emphasis for nurses and other clinicians directly involved in the MDS process.

      6. Purposes Of The RUG-III System And Its Relationship To Existing SNF Level Of Care Criteria

        According to HCFA, the RUG-III system serves "three distinct but related purposes" with respect to the regulatory certification requirements that a patient needs skilled nursing care. 63 Fed. Reg. at 26,283. First, it streamlines and simplifies the process for determining whether a beneficiary meets the statutory criteria for SNF level of care by making a per se d