Print This Page

Implementing a Standardized Post-Acute Reimbursement System

The cost of health care has been ever-increasing during the 1970's and 1980's in spite of Medicare establishment in the mid-1960's. The cost of health care had exceeded the gross national product by two digits; therefore, steps were taken to provide cost containment for healthcare diagnoses. The first attempt was controlling the major cost, which was within the hospital as new technologies allowed patients to recover from their acute illness. However, the length of stays was quite long since there was no post-acute continuum at that time.

In 1983, the DRG system became operational, however to ensure that the patients were provided a post-acute standard of care, multiple post-acute care settings were established under different reimbursement models. These systems included psych hospitals, rehab hospitals, long-term care hospitals or chronic hospitals and skilled nursing facilities. The psychiatric hospitals and rehabilitation hospitals operated under the TEFRA rate, which was established on either the first, or third year average cost based on length of stay and services that were provided for ten rehab diagnoses and certain psych diagnoses.

The skilled nursing facility was reimbursed on a cost basis without significant limitations as at that time individualized therapy services were scarce in these settings.  The long-term care hospital provisions allowed for 100 days of reimbursement based on chronic diagnoses with an average length of stay of 25 days.  These settings were to ensure that as cost containment was targeted for the acute care hospital, patients with specialized needs were still provided a continuum of services prior to being transitioned to an outpatient setting or home.

The DRG system was based on defining the patient characteristic by diagnosis and then adjusting the reimbursement based on secondary co-morbidities or complications.  The DRG defined the patient upon admission for criteria that allowed a hospital stay.  Co- morbidities were factored in on admission as well as upon discharge complications or variances were included to reimburse for the services that were provided. 

Additional provisions were made of quality assurance to provide medical stability upon discharge such that hospitals were not "dumping" the patients to maximize the reimbursement system.  These internal mechanics of quality assurance were monitored by state and federal guidelines and oversight was provided by the joint commission of accreditation of hospitals or JCAHO.

Although DRG was a dramatic shift in physician practice as well as hospital practice, the hospitals were able to survive this reimbursement model and continue to function without great impact on services being provided or any significant staff reductions. Throughout time, it was still evident that technology proceeded forward as well as physicians practices became more specialized and adjusted to the DRG limitations.

The post-acute continuum developed very rapidly by entrepreneurs who were able to maximize the reimbursement models to their benefit achieving great returns to their investment, The rehabilitation model utilized TEFRA having to incorporate the ten most common diagnoses for rehabilitation to represent, 75% of their population served within the hospital. 

The for-profit rehabilitation systems were able to manipulate the reimbursement systems for optimal revenue streams. This allowed for consolidation to the point of one major player monopolizing the rehabilitation industry.

The psychiatric hospitals became one of the fastest growing Fortune 500 companies. They were able to maximize on the benefits provided for psychiatric patients as well as alcohol and drug abuse patients. Unfortunately, these patients did not have outcomes that were sustainable and frequent readmissions were burdening the cost of health care insurance.

Additional abuses occurred with recruiting inappropriate patients and manufacturing diagnoses, which were difficult to monitor from an outside audit. The changes that were made to the psychiatric industry subsequently caused a great shift from an isolated hospital unit to incorporating a psychiatric unit within a hospital.

The long-term care hospitals are chronic hospitals, which were primarily centered in the Northeast. They were established initially for long-term patients for neuromuscular diseases, respiratory diseases, ventilators and spinal cord and head injury patients.

These hospitals were providing care for the most complex patients and through their development were able to recruit specialized physicians as well as specialized services and laboratories within their institutions to be self-sufficient. Due to the fact that they represented such a small minority of health care costs, these entities were not a significant factor in the initial reengineering process.

The skilled nursing facility market essentially was an expansion of the initial retirement home model dating back to the 1950's. Through the 1960's, patients looked upon it as a place of retirement and minimal assistance from one caretaker that would be providing the needs within a facility.

As this population to grow and medical needs increased through the 1970's and 1980's, steps were taken by Congress to ensure that proper assessments as well as the appropriate services were provided to meet the needs of these residents. In light of very few physicians, therapists, respiratory and psychiatric personnel participated in skilled nursing facilities, the reimbursement model allowed for full reimbursement based on cost of services provided.

During the late 1980's and early 1990's, the psychiatric hospitals and rehab hospitals showed significant growth as well as significant healthcare dollar utilization. The hospitals were utilizing these centers as their post-acute continuum to manage their DRG's appropriately. Unfortunately, the costs of the reimbursement models became quite high, and with the volume of patients transitioning through these centers, it added to the burden of health care cost containment.

Additional steps were taken within these entities to reengineer underutilized or unprofitable units within the hospital and convert them to individual licenses which would bring an extra source of revenue to the entity as well as share revenue gains by the before profit managing company of rehabilitation or psychiatric hospital licenses.

In the early 1990's, managed care became a prominent entity for healthcare revenue sources. As managed care tried to find the most cost effective and efficient setting, many healthcare providers felt that the skilled nursing facility license was the most diverse and non-restricted to provide most of the post-acute services and specialized services.

The routine medical post-surgical services were easily added on to a skilled nursing facility with the addition of specialized staffing, however with greater clinical expertise, specialized programs such as ventilator, dialysis, telemetry, intravenous medications, rehabilitation, head injury and spinal cord rehabilitation, TPN therapy was incorporated into these skilled nursing facility licenses. These skilled nursing facilities competed directly with the rehabilitation hospitals, and during the early and mid 1990's became the most prominent rehabilitation venue for post-acute settings.

In the mid 1990's, Medicare was being strained through the volume of patients utilizing their part A services. Congress established and enacted the Balanced Budget Act in 1995 directing for cost containment with a budget limit in July of 1998. Capitation and Medicare HMO's were options that healthcare providers could choose to manage Medicare dollars more effectively. This form of management was termed capitation where care planning and cost containment drove the healthcare resource models.

In 1998, the skilled nursing facility which were at the time the fastest growing with the highest revenue profits faced a changing reimbursement system in the form of RUG. This model was developed dating back to the 1970's where healthcare professionals monitored direct nursing times as well as therapy times for services being provided within various resident characteristics. At the time of implementation, this model seemed to hold the highest level of validity for these services.

Unfortunately, there were many undocumented and untracked resources that were being provided that were not taken into account in the development of this reimbursement model. As a result of this implementation, almost all of the for-profit had undergone bankruptcy within a couple of years. The smaller non- profit entities are functioning, only being able to provide minimal medical services and rehabilitation services for their individual residents.

Additional changes have also occurred in the rehabilitation hospital system. Instead of utilizing a cost-based model, a reimbursement model based on functional outcome was incorporated. Unfortunately, the model based on FIM measures has not solved the problem of cost containment as well as identifying the appropriate length of stay for service or services that are required for each patient characteristic. The rehabilitation model does offer a new tool, the FIM, which could be incorporated into other reimbursements models as a definer of functional outcome and intensity of rehabilitation staff that is needed.

With the restructuring of the post-acute reimbursement systems in the skilled nursing facilities and the rehabilitation hospitals, the long-term care hospitals were presented with an opportunity to evaluate the patient characteristics and costs as they relate to managing for the more chronic or long-term medically complex patient. Unfortunately, as in all aspects of life given an opportunity under a new 'license, this has created a significant growth in long-term care hospital licenses as a substitution for the skilled nursing facility licenses that were placed within freestanding hospitals.

The long-term care hospital reimbursement is very lucrative if one is able to manage the length of stay of greater than 25 days with appropriate diagnosis. Unfortunately in the early 1990's, an attempt was made to have a DRG-based system of reimbursement for long-term care hospitals, however the model was not successful at that time.

The long-term care hospitals have presented a case that they have daily physician visits which occurs in all other post-acute settings, their nursing staffing acuity is at a much higher level of 7.5 hours per patient day, and they have greater costs related to specialized equipment and laboratory uses for monitoring the patient's medical status. The long-term care hospitals as well as the specialized skilled nursing facility units are essentially identical in providing the care for the complex patient characteristics.

Unfortunately, the patients that are leaving the acute care hospitals do not have specific defining characteristics that allow placement of that patient to the most cost efficient center. The additional problem of long-term care hospital is that the reimbursement remains the same as the patient improves, however as they transition to a level of medical stability or outcome, they do not have a method of analyzing the service predictors for that level of care.

To summarize, the healthcare system prior to DRG's operated such that medical stability for all medical problems including rehabilitation and psychiatric were managed to the point of transitioning home in the acute care hospital setting. The skilled nursing facility was utilized more as a retirement assisted-living center, and long-term care hospital setting was for specialized needs within different regions for patients who had very poor prognosis but were medically complex and could not be handled long term in a higher cost environment as the acute care hospital.

The transition in 1983 with the implementation of DRG's would have been an excellent time to establish a standardized post-acute reimbursement model based on diagnosis. Unfortunately, the post-acute market became more of a boutique entity with various reimbursement models that over time have been learned and maximized to the benefit of the provider. Although we have reengineered to a degree, the post-acute license reimbursements, the goal of the unified post-acute reimbursement system is still distant.

I had been the clinical medical director of a for-profit entity in rehabilitation, skilled nursing facilities, subacute units, long-term care hospitals, outpatient therapy centers, respiratory therapy services and home health. I have been with many companies at a corporate level and was able to learn and maximize the reimbursement system based on the patients that were serviced within each license.

During the capitation, this was my easiest challenge to provide resources that served the patient's needs and maintained a viable outcome in a short length of stay for the physician groups that undertook the HMO model of Medicare and capitation. As these challenges became more complex and shorter length of stays, less reimbursement while ensuring outcomes, I reviewed the patient characteristic services that were provided and found many similarities that would yield to a post-acute reimbursement system.

My last medical director, chief medical officer position in a healthcare entity was Olympus Healthcare Corp. in 1994-1998. We had managed and owned 47 skilled nursing facilities, 15 of which were under RUG in Maine. Additionally, we had three long-term care hospitals in New England and one long-term care hospital unit within a combined licensing health center. Additional components to our organization were a pharmacy, a rehabilitation company, a respiratory therapy company as well as developing a home health entity.

At Olympus, the clinical team and I saw that although the licensing and reimbursement systems differed, the patient characteristics were homogeneous and for the majority of the patients any setting in an inpatient post-acute license would be suitable for that patient. The same characteristics and admission criteria provided for various lengths of stays within the same outcomes.

The determining factor of lengths of stays was the reimbursement models of capitation versus the traditional Medicare reimbursement in skilled nursing facility or in long-term care hospital.  Those patients who were under capitation, consistently showed greater than a 50% reduction in their length of stay without compromising outcomes.  It became obvious that physician practice could be modified to achieve the same outcome in a shorter length of stay.

Anticipating significant growth and adjustment, the goal of Olympus was to develop a predictive model of care that would provide a care map for all the post-acute diagnoses which were approximately 242 of the possible 495 DRG's.  The goal was to keep it simple, unified and not specific to one reimbursement system. The financial model was provided in parallel to the clinical model.  Costs were monitored on direct and indirect basis to ensure that appropriate real-time analysis could be done for our organization.  Nursing staff was monitored on a daily basis as well as units of rehabilitation therapy with pharmaceutical cost based on unit cost and pharmaceutical durable medical equipment cost.

Outcomes were continually being monitored for functional mobility transfers, self-care needs, bowel/bladder, grooming as well as nutrition and communication.  Primarily these are all functional benchmarks that are incorporated in the FIM model.  Additionally, to ensure medical stability for the outcome, wounds, respiratory status, psychological needs, education, clinical stability and nutrition were monitored to ensure that upon discharge the goals were met according to state and federal standards that were implemented in the acute care setting.

The goal of these outcome measures were to be consistent from the acute care hospital such that tracking could be accurate as well as sustained should they require home health services or outpatient services upon discharge from the inpatient setting. This tracking of indicators that start from the acute care hospital to the post-acute inpatient setting into home health give the best clinical information of stability, progress and outcomes.

This consistent tracking of common outcome measures should also be incorporated in a reimbursement model for healthcare in a way to avoid under representing the abilities of the patient in the transitions and then overemphasizing their progress to get the greatest change needed in order to obtain the reimbursement for the services that were provided in that setting.

The post-acute market for healthcare delivery was identified within our analysis to four levels of nursing staffing intensity. These staffing intensity levels were less than or equal to three hours per patient day, 3.1-4.75 hours per patient day for the second level, 4.76- 6.50 hours per patient day for the third highest level and finally 6.51-10 hours per patient day for the highest level of intensity. By categorizing these four intensity levels, we were able to predict the cost of nursing staffing for this type of patient characteristic.

In regards to rehabilitation therapy services and physical occupational speech and respiratory services, it was felt that the cost of these services was fairly equivalent with minimal variations. It was also felt that outside of respiratory services the traditional rehabilitation services can be divided into four categories, those being 0.5 hours per day for the lowest level, 1.5 hours for the medium level, 2.5 hours for the higher level and 3.5 hours for the highest level. The highest level would be the severe strokes, head injury and spinal cord injury patients that would require the three-level services.

At the present time, the RUG system does not define the number and type of services based on ICD-9 coding which would be more representative of the therapy needs of that patient's characteristic.

Having identified the nursing intensity level as well as the therapy level based on the ICD-9 code or DRG code, this combination would have a cost-basis that would be consistent as form of direct costs of therapy and nursing during that patient's stay. As the patient's status improved through functional improvement based on FIM, then both the nursing and therapy levels can be decreased for a more appropriate reimbursement payment system.

One of the greatest difficulties in post-acute DRG modeling is that the co-morbidities are not factored in these DRG's. Significant co-morbidities of dementia or FIM scores less than 85 are quite detrimental to the patient's progress and require greater needs and intensity of services in spite of the ICD-9 or DRG code. These aspects of dementia, incontinence, previous immobility in a wheelchair or FIM less than 85 contribute greatly to the ancillary costs as well as the durable medical equipment and lack of progress in functional physical recovery measures.

To summarize a very complex problem of co-morbidities and how they affected cost and length of stay, we found that there are three levels of co-morbidities and variances that add to these weights. The first DRG post-acute leveling would be a clean ICD-9 without any co-morbidity or variance. The second level, which would have significant co- morbidities defined on respiratory, cardiac, renal or mobility, would have a factor of 20% increase in the length of stay and cost.

The third level, or the highest degree of complexity of co-morbidities, would have cognitive deficits either preexisting or related to a new disease process, a preexisting functional FIM score of less than 85 and other more severe medical conditions that would add a 35% cost to the DRG reimbursement. Additional weight was added by ancillary cost of medication and durable medical equipment. Medication, as we owned a pharmacy, we were able determine what the true average wholesale price was and made a slight adjustment of 20% increase for processing and handling to ensure that our costs had been covered without any losses.

To summarize a DRG system and post-acute would be based on an ICD-9 primary diagnosis with secondary diagnosis to be coded as they affect co-morbidities and then the third level would be the variances or complications that occurred during hospitalization to add an added weight to the primary DRG reimbursement model. The ancillary cost of medications or pharmaceuticals and durable medical equipment can be identified by their wholesale price with a 20% adjustment based on physician and therapy recommendations.

It was important to our system that we do not interfere with physician practice management of medications other than to provide a more cost effective formula from which they were able to choose a similar medication. Curiously enough, those physicians in the model of capitation almost exclusively used all the formulary medications without any overriding of the substitution by the formulary.

The Progressive Care Path that we had developed at Olympus Healthcare was both a clinical mapping tool, a resource-tracking tool, a functional measure progression tracking tool as well as a learning tool. We identified eight areas of clinical monitoring and performance. Five of these eight categories were consistent for all medical rehab DRG paths. The first two were FIM-based based on self-care activities of daily living and the second mobility as they correlated with this rehabilitation functional measure tool.

The third was complication list, which would be a common co-morbidity to the diagnosis or post-surgical procedure. We had reviewed all of the medical literature including Harrison's Internal Medicine and other resources that helped identify the four to five most common complications for each of the diagnoses. The fourth category was a psychosocial to identify family and patient's goals, psychiatric needs and post-inpatient stay aftercare needs. The fifth category that was common to all of the paths was education of the patient and family of safety complications, medications, supplies and assistive devices that were provided by staff.

This left us with three variables that would be more specific to the DRG diagnosis.  These were identified as their primary concern based on the condition which would have included skin integrity or wound management, pain, cardiac rehabilitation, pulmonary rehabilitation, nutritional needs or bowel or bladder restoration which again utilized the rehabilitation tool, FIM.  Additional tracking was done on medical stability indicators which included laboratory tests X-Rays, vital signs and cardiac and respiratory vital signs and measures.

Therapy services were leveled or chosen of their intensity by the diagnoses. The FIM scores indicate the level of function.  For those patients that were severely impaired and required total or maximal assistance in their activities of daily living, the goal is to progress them to a functional ability of minimal assist.  This improvement of function allows therapy services to be adjusted accordingly; whereas, with maximal assistance more services and with minimal assistance less intense therapy service.

The endpoints of in-patient stay were determined by functional ability of medical stability.  When a patient reached a level of minimal assist and had the availability of a 24-hour caregiver who was capable, this was the marker of outcome for discharge.  However, when a patient was without the availability of a 24-hour caretaker who could provide for assistance, then the level of standby assistance with determinants of 150 feet with safety and some form of continence of bowel or bladder were the benchmarks for the transition to an out-patient setting.

Reviewing the literature, we found that the skilled nursing facilities and rehabilitation hospitals had very similar outcomes.  In fact, the skilled nursing facilities or subacute units were much more cost effective than the rehabilitation hospitals.  The rehabilitation hospitals showed better clinical outcomes in bowel and bladder management as well as swallowing for people with dysphasia.  Otherwise, the skilled nursing facilities were superior in their clinical effectiveness and efficiencies when measured in the mid-1990's.

The optimal resource containment was maintained by capitation systems, which showed similar outcomes in a very short duration of length of stays while maintaining medical stability. In capitation, physicians either made rounds daily or at least twice a day and were allowed to discharge late in the evenings, which allowed for a reduction of at least one-day length of stay.

In conjunction to the progressive care path system, there was a quality monitoring system based on clinical indicators. These clinical indicators were established based on the type of setting whether it was chronic or long-term care hospital versus a skilled nursing facility or a rehabilitation model. The state survey process, the joint commission accreditation process as well as the commission on accreditation of rehab facilities, indicators were utilized within an indicator system.  Literature was reviewed to establish national statistics on incidents of certain outliers. These were used as thresholds not to be exceeded by our own internal institutions.

This quality monitoring system can be simplified to a post-acute system incorporating the entire medical, surgical and rehabilitation measures of medical stability and quality during the inpatient stay.

Additionally, the same indicators that were used to track the inpatient stays in the hospital and the post-acute system can be translated into home health system such that universal base of clinical indicators could be placed for the duration of the DRG for each patient characteristic. In the future, this tracking mechanism of keeping common indicators is much more viable in data collection and review in order to provide efficiencies and clinical effectiveness when streamlining the DRG timeline.

Should you have any further questions regarding this article, please direct your questions or comments to "Ask the Doctor" section.

Copyright © 2004 - 2008 Taras V. Kochno, M.D.  All Rights Reserved
Board Certified in Physical Medicine and Rehabilitation

HOME