In its landmark review of the causes and consequences of disability, the Institute of Medicine (IOM) confirmed the importance of rehabilitation in restoring and maintaining quality of life, especially among older adults.1 In an effort to meet the rehabilitation needs of this population, Medicare has covered outpatient therapy services since the enactment of Title XVIII of the Social Security Act in 1965. In 2006, 4.4 million individuals received outpatient therapy services (physical therapy, occupational therapy, and speech and language pathology), representing 9.7% of the 45.5 million beneficiaries receiving Medicare Part B services.2 In its follow-up report, The Future of Disability in America,3 the IOM further concluded that antiquated policies, which would include those policies undergirding current payment methodology for outpatient rehabilitation services, may delay or interfere with the individual's eventual functioning. In order to achieve payment policies for rehabilitation services that promote quality of life while prudently allocating resources, the concepts underlying these policies should receive the same level of reflection that is now routinely applied to the scientific concepts that support best clinical practice. The purpose of this article is to present and discuss a conceptual basis for an alternative payment system for outpatient therapy services based on the therapist's judgment of severity of the patient's condition and expected intensity of treatment for that condition.


Medicare coverage and payment policies are based on the setting in which they are provided; therefore, policies among the various settings may differ significantly. Medicare Part B, which covers outpatient therapy, requires that the services be (1) provided by a skilled professional as defined in the Center for Medicare & Medicaid Services' (CMS) qualified provider language, (2) medically necessary (ie, deemed appropriate and effective for the patient's condition), and (3) reasonable in terms of frequency and duration. Services provided to Medicare beneficiaries are classified and reported to CMS according to the Healthcare Common Procedure Coding System, based on the American Medical Association's (AMA) Current Procedural Terminology (CPT).4 The CPT comprises a set of codes, descriptions, and guidelines intended to describe procedures and services performed by physical therapists, physicians, and other health care professionals when billing for services to third parties. For example, there are specific CPT codes for physical therapy evaluation, therapeutic exercise, manual therapy, and ultrasound. The value of each CPT code is based on the Resource-Based Relative Value Scale (RBRVS), a methodology that uses weights, called relative value units, to represent the relative costliness of the inputs used to provide the services. These inputs are defined by AMA as clinician work, practice expenses, and professional liability insurance. Most important to the concerns of this article, RVRBS methodology calculates a work value by considering 5 factors: mental effort and judgment, technical skill, physical effort, psychological stress on the provider, and time (Appendix).5 Medicare determines payment for each therapy service described by a specific CPT code using the Physician Fee Schedule, which itself is based in part on the RBRVS and geographically adjusted cost estimate.

Currently, physical therapists are confined in CPT to using only 2 “evaluation” codes and a larger array of codes for specific procedures. Moreover, the majority of the more “intense” procedures (eg, a cervical manipulation coded in CPT as “manual therapy”) must meet specific “time” requirements (ie, 15 minutes of “service” for a procedure that typically has a brief duration) in order to be considered medically necessary and meet the threshold for payment. In comparison with accounting for each procedure or service separately, physicians and some other nonphysician providers such as nurse practitioners and physician assistants are allowed to use evaluation and management (E&M) codes that do not require denotation of every specific procedure performed or service provided during each patient visit. Each E&M code describes a different kind of provider-patient encounter and accounts for patient type (new or established), setting of the service, and level of E&M service performed. The level of the E&M service is determined by patient history, examination, medical decision making, counseling, coordination of care, nature of the presenting problem, and time.6

It is our belief that the prevailing metric for “quality” in rehabilitative care often has been imprecisely equated to the time spent with the patient in performing a procedure. Although time may be the most easily quantifiable factor, it may be the least appropriate proxy measure for judging the appropriateness of the care received. Coding and payment for therapy services under CPT and RVRBS, particularly as it has been developed in the 97000 series (Physical Medicine and Rehabilitation), adequately account for the clinical work described by the second through fifth factors. However, we contend that the current application of CPT to therapy services does not fully account for the defining characteristic of contemporary professional practice among rehabilitation clinicians, which is delineated by the first factor (ie, mental effort and judgment). Moreover, this inadequate conceptualization of practice hinders the recognition of how the mental effort and judgment of therapists affect the design, implementation, duration, and ultimately, the costs of the treatment plan across the episode of care. In contrast, we believe that a broader conceptualization of work that emphasizes continuous examination and the multiple components of clinical decision making and patient care management will facilitate the determination of medical necessity and appropriateness of care.

Therapy services represented $4.07 billion in 2006, with physical therapy accounting for 75% of submitted Medicare claims.2 For more than a decade, Congress and CMS have expended considerable effort to develop various payment policies and methods that might control the cost of services while enhancing the quality of care. Specific to therapy services, Medicare has experimented with instituting a financial cap, implementing exceptions to the cap, limiting appropriate payment for specific procedures or combinations of procedures, and establishing a system to measure therapist quality by enabling some physical therapy providers to participate in the Physician Quality Reporting System.79

Although great effort has been demonstrated to develop and institute these polices, the effectiveness of such initiatives either to control costs or to sustain quality is still being debated by providers, patients, and policy makers. The CMS has recognized that the current system of payment for therapy services under the Medicare program requires change. As described in its own report, CMS “has articulated a vision for health care quality—the right care for every person every time…. Medicare's current fee-for-service payment system, which pays on the basis of quantity and consumption of resources, does not support this vision for quality health care.”7

In order to change the system, CMS has funded several projects to propose payment policy changes for therapy services. The Developing Outpatient Therapy Payment Alternatives (DOTPA) project was awarded to RTI International in 2007 to “identify, collect, and analyze therapy-related information tied to beneficiary need and the effectiveness of outpatient therapy services…with an ultimate goal to develop payment method alternatives to the current financial cap on outpatient therapy services.”10 A major focus of this long-term initiative is to develop and test assessment tools and methods that will ensure appropriate therapy services are provided to beneficiaries efficiently.

A second project, Short Term Alternatives for Therapy Services (STATS), was awarded to Computer Sciences Corporation in 2008 to develop alternatives to outpatient therapy caps in the short term that will encourage payment for medically necessary services. For this project, “short term” was defined as methods “that can reasonably be expected to be implemented within 2–3 years from approval by CMS.”2 The focus of this project was to develop alternatives that are “clinically appropriate,” “operationally feasible,” and effective in ensuring appropriate access to therapy for beneficiaries who need services, yet limit unnecessary expenditures. The major focus of this project was to develop alternatives within the present Medicare Physician Fee Schedule and claims system and enact change through policy revision and coding changes.

Although these efforts to review and recommend improvements to the present system are warranted, we believe that these initiatives focus only on understanding the quantity of services and neglect the clinical processes that drive the provision of therapy services. In brief, our contention is that proposals for alternative payment that do not fully account for the processes of clinical diagnostic reasoning and skillful implementation of the plan for intervention are essentially inadequate as alternatives to the present system. Specifically, we propose that a more thorough understanding of how a therapist determines the need for services, the factors that are considered in the decision, and the relationship between determining the need for therapy based on the severity of the condition to be treated and the concomitant intensity of the treatment provided will enhance any alternative payment methodology. Thus, we believe that a payment methodology that resonates with the decision-making processes that lead up to and surround the implementation of a treatment procedure is more likely to support providing the right care to the right patient at the right time at the right cost.

Severity-Intensity Concept

The acceptability of any payment system is judged on its ability to provide needed care to beneficiaries in a timely and effective manner and to pay providers for such services fairly. Currently, in order to be paid for therapy services under Medicare Part B, the provider must demonstrate the “medical necessity” of the care.46 “Medical necessity” is determined by information included on the claims form, which contains very little clinically related information besides medical diagnosis. Thus, medical necessity, by default, is based primarily on the medical diagnosis or, more likely in the case of older adults, diagnoses. However, medical diagnosis alone reveals very little about the need for therapy.

We believe that the judgment determining medical necessity should be based on the interplay of 4 factors: the medical conditions of the patient, the physical impairments resulting from these conditions, the patient's ability to function, and an assessment of the specific sociocultural and environmental factors that enable the individual to participate in his or her various societal roles, particularly as the last factors are described by the International Classification of Functioning, Disability and Health (ICF).11 These 4 factors, taken together, underlie the provider's judgment about the “severity” of the patient's condition and the need for physical therapy services by aggregating data into a cohesive evaluation of the magnitude of the difference between optimal and actual examination findings across physical, psychological, and social domains of function and health status. In some instances, a medical diagnosis alone, such as complete spinal cord lesion at C6, may be sufficient, in and of itself, to affect the judgment of severity. In other cases, the interactions among various combinations of these factors may prove to be more decisive in how a therapist determines patient severity. For example, 2 individuals with the medical diagnosis of osteoarthritis could differ significantly in severity level based on the number and impact of the resulting impairments, the inability to function in their daily activities, and their ability to meet societal expectations. A 45-year-old construction worker with severe erosions of the small joints of the foot that greatly impede weight-bearing functional activities, alter gait, and diminish work activities presents a substantially different functional challenge than a 45-year-old computer programmer with similar impairments who teleworks. Yet, in all of these instances, the therapist's judgment of severity, rooted in good clinical decision making, is essential to designing and implementing the plan of care. This critical notion of how the therapist judges the aggregate impact of the patient's presentation on function and health is essentially missing in the present approach to determining medical necessity.

Payment for therapy services under the Medicare program further requires that the provider justify the implementation of a plan of care.46 The therapist must develop the components of this plan of care with sufficient intensity to meet the expectation that such treatment will lead to realistic improvement in the patient's condition in a reasonable period of time. Establishing “intensity” involves a series of judgments that a therapist must make regarding the particular combination of procedures to implement: how much of each procedure to deliver at any one encounter, the administration parameters of the procedure itself, and the optimal frequency of the procedure over the episode of care. Again, clinical judgment by the therapist is essential in determining the intensity of care.

The present payment perspective is based largely on the medical model of an acute episode that presumes the relationship between the severity of a patient's condition and the intensity of the treatment received would be direct and linear. Simplistically, the “sickest” patients receive the most intense and immediate efforts to restore health or slow disease progression.

It is not self-evident that this payment model built around acute illness applies equally to outpatient rehabilitation services for which the relationship between the severity of the condition and the intensity of the treatment may not be linear. For example, the initial intensity of care for a patient with a complex presentation of medical comorbidities and impairments such as might be seen in an older adult with metabolic syndrome may be fraught with greater risk due to the severity of the patient's condition and require greater mental effort and judgment during implementation of treatment. In fact, an appropriate clinical strategy may require a graduated increase in intensity of intervention to achieve a desired outcome only as the initial clinical presentation resolves. For example, a patient with cervical whiplash and comorbid osteoporosis referred shortly after a motor vehicle accident and receiving services in the outpatient setting would likely be judged by the therapist as very severely affected shortly after emergency room discharge, but may not immediately receive the most intense treatment. Therefore, one would anticipate that services would increase over time as the immediate effects of injury such as pain and swelling decrease while motion and activity tolerance increase. In fact, as severity subsides, intensity of treatment would increase. Finally, as the patient continues to improve functionally, a tapering of services would be likely as discharge approaches. In contrast, a patient with cervical whiplash injury and without other skeletal comorbidities would be judged as less “severe” and might do well with a moderately intense intervention schedule implemented early in the episode of care, but then be tapered to interventions requiring little supervision as his or her improving condition poses inherently less risk to his or her safety over time. Thus, although both patients carry a medical diagnosis of whiplash, neither exhibits the same severity, nor would similarly intense treatment programs be appropriate.

A payment system based primarily on the quantity of procedures with little association to the therapist's assessments of the severity of the patient's condition, the mental effort required to provide the care, the intensity of the intervention, and its inherent risks would undervalue the first case, where appropriate caution places extra burden on the provider, and may overvalue the second case, where a less severe clinical presentation allows for more treatment. Understanding severity and intensity as drivers of effort and cost is essential if we are to create a system that ensures both the right care of the right quality to the right person and appropriate allocation of resources.

The concepts of severity and intensity are not unique concepts for the study of health care utilization,12 although this is the first application to rehabilitation. Contemporary physical therapist practice is distinguished by its emphasis on the independent judgments of physical therapists regarding the patient's condition and the design of the plan of care to address that condition. The incorporation of the concepts of severity and intensity into an alternative payment system for rehabilitation would be more completely aligned with actual practice if a conceptually broader array of descriptors were available in CPT and were accessible to physical therapists as part of the alternative payment system.

The severity-intensity concept will provide additional utility as well for legislative and regulatory efforts to establish the appropriate cost of care based primarily on outcomes. It is our belief that, even if adjusted for other factors, proposals calling for an outcomes-based system fail to adequately account for how complex clinical judgments about the severity of a patient's condition help to frame the selection of interventions that contribute to variations in an episode's duration and cost. In order to achieve beneficial outcomes, physical therapists must first make accurate judgments to determine the severity of the patient's condition and then make appropriate decisions regarding which interventions to use based on this patient assessment. If the state of rehabilitation science allowed us to know with certainty how patients with like conditions should be clustered in order to predict what outcomes could be routinely achieved after receiving appropriate doses of an intervention, outcomes-oriented approaches to payment might be proven more fruitful than they have been to date. We believe that a system responsive to a therapist's ability to ascertain severity correctly and subsequently match these judgments to an appropriate intensity of intervention is a more suitable starting point to develop “alternative” payment policy. This approach can be implemented as we wait for more robust population-based data on the likely outcomes of treatment across the wide spectrum of conditions for which physical therapist intervention is warranted, as well as professional consensus on what is appropriate intervention.

Directions for Future Research

Severity and intensity can be challenging concepts to grasp initially and to define operationally. Our specification of the concept, clearly, is only the first stage of efforts to further its adoption by policy makers. In order for the severity-intensity concept to be accepted as the basis for an alternative payment system, its elements require additional verification and refinement. Specifically, the hypothesis that therapists possess the ability to consistently make clinical judgments regarding the severity of a patient's condition and the intensity of the intervention that leads to appropriate desired outcome must be conclusively demonstrated if the concept is to become the basis of a payment methodology. These judgments would have to be consistently made by therapists submitting charges for services to the Medicare payment system and accommodate the normal variability of judgment as a function of the clinician's education, expertise, and experience. A series of studies will be needed to clarify and confirm the criteria that best allow physical therapists to accurately and consistently stratify the severity of a patient's condition and the intensity of the treatment provided for that condition. The RVRBS was developed by such a research program,13 and the provisional operationalization we have given to the concept is only the first step toward establishing a comparable research plan.14

Potential Applications of the Concept

If it can be demonstrated that physical therapists possess a common understanding of severity, we believe that they also will be able to achieve consensus on what is the appropriate treatment intensity for a given level of severity of a particular condition. Although we believe that outcome-oriented approaches to payment may be premature, the severity-intensity model could serve as an interim step toward using outcomes as one component of a payment methodology in the long term. If we imagine a simple system that would allow for a classification of a patient according to 3 levels of severity (ie, low, moderate, and high) and a similar tri-level assessment of intensity such as depicted in the Figure, we could establish a rudimentary coding scheme. Specifically, patients could be stratified by severity of condition and intensity of service utilization and thus more appropriately compared. In this way, we would be assured that clusters of similar patients could be compared for similarities in treatment and enable the identification of outliers. Coupled with data derived from reliable and valid measures of outcome matched to the patient's condition, we would be able assess whether the outcome achieved was optimal for a specific level of severity. It is not our contention that such a system would lower costs or lead to better payment for services. However, such a system would more likely support providing the right care to the right person at the right time. Thus, a particular therapist's remuneration would depend on types of patients seen and the patterns of care provided, not merely on the volume of procedures performed, as is currently the case.


A schematic representation of coding based on severity-intensity stratification.

In conclusion, we believe our approach to developing a system that decreases unwarranted variation in clinical practice by promoting service delivery that is concordant with evidence-based expectations about treatment for similar patients and increases provider accountability for performance could be initiated in the near future. This system for outpatient therapy services would be characterized by an enhanced emphasis on clinicians' common understanding of the severity of a patient's condition and the intensity of treatment necessary to address the requirements of each individual. We believe the concept is sufficiently robust to justify advocating for it as the basis for an alternative payment methodology.


Components of Clinician Work Used to Value Service Under the Resource-Based Relative Value Scale5

Mental effort and judgment necessary with respect to the amount of clinical data that needs to be considered, the fund of knowledge required, the range of possible decisions, the number of factors considered in making a decision, and the degree of complexity of the interaction of these factors.

Technical skill required with respect to knowledge, training, and actual experience necessary to perform the service.

Physical effort can be compared by dividing services into tasks and making the direct comparison of tasks. In making the comparison, it is necessary to show that the differences in physical effort are not reflected accurately by differences in the time involved; if they are, considerations of physical effort amount to double counting of physician work in the service.

Psychological stress—2 kinds of psychological stress usually are associated with physician work. The first is the pressure involved when the outcome is heavily dependent upon skill and judgment and an adverse outcome has serious consequences. The second is related to unpleasant conditions connected with the work that are not affected by skill or judgment. These circumstances would include situations with high rates of mortality or morbidity regardless of the physician's skill or judgment, difficult patients or families, or physician physical discomfort. Of the 2 forms of stress, only the former is fully accepted as an aspect of work; many consider the latter to be a highly variable function of physician personality.

Time it takes to perform the service.


  • The authors acknowledge the expertise and assistance given by Jim Nugent, Gayle Lee, JD, Courtney Merritt, and Michael LaValley, PhD, in the preparation of the manuscript.

  • Received February 9, 2011.
  • Accepted June 19, 2011.


View Abstract