Introduction

“Cancer rehabilitation is integrated throughout the oncology care continuum and delivered by trained rehabilitation professionals who have it within their scope of practice to diagnose and treat patients’ physical, cognitive, and functional impairments in an effort to maintain or restore function, reduce symptom burden, maximize independence, and improve quality of life in this medically complex population.” [1]

As cancer transitions from a terminal disease to chronic medical condition, so too must our viewpoints regarding rehabilitation as it fits into the continuum of care. Survivors are more likely to experience physical and psychological impairments as a result of cancer and its treatment, and these in turn can affect function and quality of life (QOL) [2]. Clinical gaps still exist for individuals with significant medical acuity and functional impairment that cannot be met in the outpatient setting. The cause for such impairments is often complex and requires a coordinated approach to provide the greatest opportunity for stabilization and improvement. For select individuals, the inpatient setting may be appropriate, as it allows for the delivery of specialized rehabilitation services while coordinating medical and nursing care in a supervised setting.

The History of Inpatient Rehabilitation Care

Most of rehabilitation care is delivered in the post-acute care (PAC) sector, which is comprised of home health agencies (HH), skilled nursing facilities (SNF), inpatient rehabilitation facilities (IRF), and long-term care hospitals (LTCH). Whereas HH is typically provided in the home setting, inpatient rehabilitation is delivered at IRF, LTCH, or SNF.

IRF

IRF provides a unique setting for individuals with catastrophic illness and injuries to focus on their physical and functional recovery. In 1977, the Health Care Financing Administration (HCFA) began standardizing the delivery of programs for the Centers for Medicare & Medicaid Services (CMS). To distinguish rehabilitation units from medical-surgical units, several admission criteria were established to deem a patient appropriate. The three primary components included medical stability, expectation of significant functional improvement over a reasonable period, and ability to tolerate 3 h of therapy daily. Although not considered a requirement for admission, 10 diagnoses were outlined as most representative for inpatient rehabilitation (Table 1) [3].

Table 1 Ten diagnoses outlined as most representative for inpatient rehabilitation
Full size table

However, expenses and costs to Medicare increased significantly over the next several decades. As a result, the HCFA created the Prospective Payment System based on Diagnostic Related Groups (PPS-DRG). Instead of paying hospitals on a per-diem basis, they would now be paid in one lump sum for each diagnosis treated. Hospitals and institutions were then responsible for the most efficient management of those funds. Rehabilitation was initially excluded from the PPS-DRG, as it did not follow the typical medical-surgical inpatient model. However, the initial 10 conditions were converted into an admission requirement, such that 75% of all the admissions to IRF should consist of these diagnoses [3].

With a separation in payment structures between medical-surgical units and IRF, hospital systems attempted to maximize reimbursement in both settings and thus developed their own IRFs to more quickly transition out of the acute care setting. Inpatient beds grew from 14,000 in 1985 to 41,000 by 2000, and total number of IRFs more than doubled from 484 to 1048, respectively. Growth led to increased Medicare rehabilitation expenditures, exceeding US$8 billion. CMS began enforcing the 75% rule and decertified programs that could not remain compliant. The total number of discharges from IRF decreased from 510,000 in 2004 to 412,000 in 2006. Given the significant changes in patient access, the rule was permanently adjusted to a 60% in 2007 [3]. Medically complex conditions, such as cardiopulmonary and cancer, were still considered “non-compliant.” However, as a result of cancer and its treatment, several oncological diagnoses could still be coded as compliant, such as spinal cord injury from metastases, non-traumatic brain injury secondary to glioma, and amputation due to sarcoma resection.

SNF

With the creation of the Social Security Act of 1935, the first federal state public assistance program, called the Old Age Assistance (OAA) program, was established. Utilization of funds stimulated the growth of voluntary and proprietary nursing home to approximately 9000 by 1954. These facilities were categorized as skilled nursing or personal care homes with skilled nursing facilities. Further federal funds were dedicated for the construction of SNF sponsored by not-for-profit organizations. These facilities were expected to meet specific definitions and hospital-like building standards. As a result, payments for the OAA increased from US$35.9 million in 1950 to US$280.3 million in 1960. The program was expanded to cover those that were deemed “medically needy,” and by 1965, the total cost was US$1.3 billion per year for 300,000 beneficiaries [4].

Given the large amounts of funding provided to SNFs, quality of care (QOC) was becoming more of a focus. In the 1950s, most nursing homes were found to have low standards of service and relatively untrained personnel. With the advent of CMS in 1965, additional funding was provided for individuals requiring post-hospital convalescence. Medicaid specifically paid for skilled nursing services at extended care facilities (ECF), but compliance with health and safety standards remained difficult. In 1972, the federal government provided funding to support state survey and certification activities and develop a set of uniform SNF provisions for both Medicare and Medicaid. Federal regional offices were established to help enforce long-term care standards for SNFs receiving Medicare and Medicaid funding [4].

After the Institute of Medicine (IOM) report to the HCFA in 1986, an expert panel reached consensus regarding the need for federal regulation in the areas of patient rights, QOC, and QOL [5]. The HCFA clarified guidelines to ensure requirements for coverages were clearer [6]. In 1988, legislation improved access to SNF care. Utilization jumped fourfold and continued to increase through the 1990s [7]. With the adoption of a PPS system for SNF in 1998, the number of Medicare residents decreased at SNF; however, rehabilitative services and therapy minutes increased with constant length of stay (LOS). Medicare volumes began exceeding levels seen prior to the PPS system [8].

LTCH

In 1984, the acute hospital PPS was established as a measure to control costs for acute care hospitalization. LTCHs were developed to provide hospital level care for patients with significant medical complexity. It was initially considered a setting for ventilator weaning, but with the specific requirement of maintaining an average LOS greater than 25 days [9]. Although LTCHs represented a small portion of Medicare spending at their inception, they experienced significant expansion between 1988 and 1996. The average annual growth rate was 31%, and expenses increased from US$200 million to US$1.7 billion annually [10]. Even with the development of a PPS system to control expenditure, LTCH still represents a significant portion of PAC cost. Although evidence does suggest benefit for select patients, data is limited regarding the overall value based on well-defined outcome criteria [11].

The Current State of Inpatient Cancer Rehabilitation Care

Although initially designed as a system to help contain cost, the PPS led to a significant amount of growth in PAC. According to the Medicare Payment Advisory Commission (MedPAC), PAC spending totaled US$60.3 billion in 2015, with SNF, IRF, and LTCH accounting for approximately US$41.9 billion (US$29.6 billion, US$7.5 billion, and US$4.8 billion, respectively). With increased enrollment and expansion of Medicare Advantage programs, fee-for-service (FFS) expenses have slowed. Although IRF has experienced slight increases in spending, PAC overall has remained constant since 2011 [12]. In the current framework, CMS has used regulations to ensure patients receive the appropriate level of care based on medical necessity and functional need. Inpatient rehabilitation admission and continued stay are very closely associated with the regulatory environment.

IRF

Reasonable and medically necessary criteria are the primary considerations for Medicare beneficiaries to receive care at the IRF level (Table 2). Although some of the expectations are clearly outlined, such as the ability to participate in 3 h of therapy per day for 5 days per week, others are more open to interpretation by the rehabilitation provider. One of the difficulties is balancing the need for medical complexity, associated with frequent physician rounding and management, with ability to actively participate in interdisciplinary therapy. Furthermore, several documentation requirements have been put in place to ensure that care cannot be provided at a less intensive setting, such as SNF [13]. Included in this are preadmission assessments, a reassessment within 24 h of admission, and a summative analysis of the patient condition, prognosis, and treatment plan within 72 h of the IRF admission.

Table 2 Primary considerations for Medicare beneficiaries to receive care at the IRF level
Full size table

Several variables can influence payment per case in IRF, including the setting by which care is provided. Considerations are made for whether a facility is hospital based versus free standing, or rural versus urban. The total number of beds within IRF and occupancy rates may also impact overall reimbursement. The average payment per case in 2015 was US$19,116. Discharge to community rates are becoming an increasingly important measure of outcome, and in 2015, 76.0% of the individuals discharged directly to the community. Only 2.4% of the patients experienced potentially avoidable re-hospitalization during the IRF stay and 4.2% potentially avoidable re-hospitalizations during the 30 days after discharge from IRF. The overall average length of stay (LOS) was 12.7 days [14]. Careful evaluation of data as it relates to healthcare outcomes, inpatient utilization, and overall costs remains a focus for IRF and is currently being integrated into reimbursement structures. Current areas of assessment include new or worsening pressure ulcers after admission, measures related to all cause unplanned admissions 30 days after discharge from IRF, and hospital onset Clostridium difficile infection rates [15]. New measures are developed, with appropriate input from stakeholders, based on perceived need and importance by CMS.

As an additional measure of expense management, CMS has also initiated an auditing program to review perceived overpayment to acute and PAC providers [16]. The Medicare Recovery Audit Program was established nationally in 2010 [17]. With the use of Recovery Audit Contractors (RAC), IRF admissions are reviewed to determine whether payments were appropriate based on patient need, level of care, compliance with provider documentation, and justification for continued inpatient services. Data supporting overpayment is controversial. Although it initially showed a threefold increase in RAC overpayment determinations, a significant amount of cases contesting denials have been overturned in favor of hospital systems. However, because denials require of immediate repayment, and appeals are both lengthy and expensive, many systems have opted to negotiate settlements or rebill under Medicare Part B programs [18].

SNF

Similar to IRF, SNF is another setting that is frequently used for inpatient rehabilitation care. However, the justification for admission to SNF rehab differs from that of IRF. Typically, individuals participating in a SNF program receive a lower intensity of medical management and therapy intervention. For the traditional Medicare program, individuals must meet a minimum 3-day hospitalization in acute care that is not considered to be an observation stay. The first 20 days of SNF care are typically covered under Medicare part A benefits, and partial coverage is also available for days 21 to 100 [19]. The conditions for participation at SNF and care provided are somewhat different than those for IRF; however, many can be applied for individuals with an oncological diagnosis (Table 3) [20].

Table 3 Conditions for participation at SNF
Full size table

Although therapy delivery may be less intensive than IRF, it is still a substantial portion of the SNF experience. Typical minutes at SNF range from 60 to 90 per day. In specific cases, an increased intensity can be provided. For example, 51% of the Very High Rehabilitation categories received 500–510 min per week, whereas 65% of the Ultra-High Rehabilitation received 720 and 730 min. Total Medicare payments to SNF in 2013 totaled almost US$27 billion [21]. Costs were somewhat less than those at IRF, with a median payment per day and per stay of US$463 and US$18,361, respectively. It remains a significant component of PAC utilization and represents the disposition for 20% of all Medicare beneficiaries discharged from acute care [22]. Like IRF, SNF is currently participating in Medicare quality reporting programs (QRP). Measures being assessed include new or worsening pressures ulcers during SNF stay, patients with one or more falls and associated major injury, and percentage of patients with admission and discharge functional assessments and care plans [23].

LTCH

The focus of LTCH is the treatment of complex medical conditions, such as oncological diagnoses, where the expected average LOS is at least 25 days [9]. Although therapy is available, its delivery is dependent on the ability of a patient to participate in context of her or his medical condition. Therapy minutes are usually less than those provided at IRF. For facilities not participating in site neutral payments, several criteria need to be met to justify FFS payments (Table 4) [24]. Overall cost of LTCH is typically the highest within PAC, and average Medicare payment per case in 2015 was US$40,718 with total payments reaching US$5.3 billion. The average LOS was 26.6 days for 116,088 beneficiaries that utilized these services [25].

Table 4 Criteria for Admission to LTCH Exclusive of Site Neutral Payments
Full size table

Outcomes for cancer patients at LTCH have not been well studied, and it is often difficult to predict which patients will be admitted to LTCH after acute hospitalization [26]. Regulatory criteria place additional complexity on diagnoses amenable to LTCH care. Data pertaining to functional outcomes is lacking due to high medical acuity and inability of some to tolerate rehabilitation services. Much like IRF and SNF, LTCH is subject to quality reporting measures, such as new or worsening pressure ulcers during the stay, change in mobility for patients requiring ventilator support, and admission and discharge functional assessments and associated care plans [27].

Impact Act of 2014

With increased federal oversight and regulation, improvements have been made regarding management of the healthcare environment. However, the costs associated with care delivery are still substantial. In 2015, total healthcare expenditure totaled US$3.2 trillion. It represented 17.8% of the total gross domestic product of the USA. A significant proportion, 32.3%, represented care provided in the hospital setting [28].

With passage of the Affordable Care Act in 2010, a renewed focus was placed on access, affordability, and QOC. The law was the first step in transformation of healthcare payment systems, with approximately 30% of the traditional Medicare payments flowing through alternative payment models (APM), such as bundled payments [29]. In association with CMS, the National Quality Strategy (NQF) uses proven interventions to improve population health and reduce the cost of quality health care [30]. Both have helped to develop the QRP for inpatient rehabilitation settings.

Standardization of outcome measures and cost have been difficult to achieve across the PAC setting. Each specific section had been operating in silos, both from a reimbursement and from an outcome perspective. The Improving Medicare Post-Acute Care Transformation (IMPACT) Act was developed so that patient assessment data could be compared with regard to quality measures and patient assessment instrument categories. Reporting timetables were created to ensure that all sectors of PAC remained compliant with the QRP. The IMPACT Act addresses assessments in terms of quality, resource management, and general assessment categories (Table 5). One key metric evaluated will be the total estimated Medicare spending per beneficiary, which for the first time will use cost as a measure of efficiency of care over all PAC sectors [31]. Reimbursements that were traditionally FFS will be linked to value based initiatives such that payments will be tied to both the quantity and quality of services provided. As it continues to evolve, the IMPACT Act may provide a unique opportunity to address the needs of specific patient populations in the rehabilitation setting.

Table 5 Outcome measures for the IMPACT Act
Full size table

Future for Inpatient Cancer Care

Efficiency and effectiveness, along with utilization of resource, are becoming a primary focus for health care. In the current regulatory environment, inpatient rehabilitation has been an underutilized service for the cancer patient. However, it provides a unique opportunity for the medical supervision necessary to address oncological complexity. Coordination of care between medical, oncological, and rehabilitation service lines allows for optimal timing of treatment while improving performance status. From a rehabilitation perspective, the current utilization and reimbursement frameworks do not clearly outline cost expectation and resource needs to care for different oncology diagnoses and causes some difficulty for rehabilitation institutions to appropriately code and subsequently manage staffing and medical supplies necessary for this population. However, with further innovation into payment structures and care delivery, many of the previously described regulations may be adjusted within APMs. This may allow oncology diagnoses to be managed throughout the entire spectrum rather than in siloed interactions. As such, the ability to manage resources may allow for appropriate care at the appropriate time to improve performance status, decreases healthcare utilization, and reduces the episodic cost of care.

With the creation of the IMPACT Act, the first steps to standardizing outcome measures in PAC are being implemented. However, further work is necessary to understand the measures that are most important to both oncology and rehabilitation service. Continued training and education will be necessary to provide safe and effective care that accounts for oncological complexity and rehabilitation precautions. However, standards are still evolving for different disciplines. In the case of physiatry, a majority of residency program directors felt that training opportunities in cancer rehabilitation were only average and that residents did not receive adequate exposure to cancer during their residency [32]. Future development of physician fellowships, as well as therapy residencies, may help to address this gap. By understanding the components of care most important to regulators and payers, as well providers and institutions, cancer rehabilitation will have the opportunity to develop standardized approaches to clinical care, outcome measures, and reimbursement.

Cost continues to be a major focus in the healthcare landscape, as it relates to value of care delivered. For any value proposition, a balance must be achieved between treatment benefits, QOC, and economic costs. Outcome measures and standardization of care are important variables that can influence this equation. However, the actual cost of care remains a significant component. Several factors can affect the overall expense associated with cancer treatment, including drug prices, acute inpatient hospitalizations during the last months of life, or diagnostic imaging. With payment reform, opportunities may exist to control cost in context of QOC [33].

With the creation of the Bundled Payment for Care Improvement (BPCI) Initiative, CMS is attempting to reduce expenditures while preserving or enhancing QOC. Although several models are being piloted, all are based on a fundamental concept that episodes of care through PAC would be addressed with a single payment managed by providers. If the costs exceed the predetermined payment, the institution would repay CMS the difference, and if the cost was lower, providers would receive the differential. BPCI programs include amputation, major joint replacement of the lower extremity, and stroke [34]. In select populations, such as lower extremity joint replacement episodes, Medicare payments declined more for those participating in BPCI participating hospitals than in comparison hospitals, with no significant change in quality outcomes [35]. Although there was a 5% increase in spending overall for joint replacement episodes, hospital savings and reduced Medicare expenditure were noted for BPCI when PAC spending was decreased [36].

Similar to BPCI, the Oncology Care Model (OCM) aims to provide higher quality and more coordinated care at the same or lower price to Medicare. It is a bundled payment system for oncology patients that incorporates practices, payers, and patients to manage episodes of care. It allows for either a Monthly Enhanced Oncology Services payment to help manage and coordinate the care for Medicare beneficiaries or performance-based payments incentivizing programs to lower the total cost while improving care for beneficiaries. Although each episode begins with outpatient chemotherapy, all services required over the next 6-month period would be managed with the bundled payment [37]. Given the common need for hospitalization during chemotherapy treatment, inpatient rehabilitation facilities may provide lower cost options to deliver inpatient care. With specified therapy services and improved performance status, inpatient interventions may lead to decreased future healthcare utilization and thus be an effective measure incorporated into the bundle.

Conclusions

Inpatient rehabilitation for oncology provides a unique opportunity for comprehensive medical care while addressing performance status. Each setting has specific considerations regarding appropriateness for admission, requirements during the inpatient stay, and reimbursement structures. However, with understanding of the regulatory and legislative environment, opportunities exist to integrate this population into currently existing rehabilitation settings. As the healthcare environment continues to innovate with novel care delivery, utilization of inpatient rehabilitation may be valuable to lower the cost of inpatient care, provide services that improve performance status, and decrease future healthcare utilization after medical and functional optimization.