Achieving Pay-for-Performance in Outpatient Practice through Measurement of Functional Health Outcomes

Author(s): Carl V. Granger, MD, Lynne M. Adamczyk, RN, BSN, MBA

Originally published:09/11/2015

Last updated:09/11/2015

1. OVERVIEW AND DESCRIPTION

The value of rehabilitation is in the outcomes.
In order to be manageable, outcomes must be measurable.

The American Academy of Physical Medicine and Rehabilitation’s 2014 handbook states, “Healthcare is trending from a fee-for-service model to a pay-for-performance system that will be challenging for physicians to adopt.” It goes on to say “the Academy will continue to strongly advocate for the development and adoption of appropriate functional outcome measures that show the true value physiatrists bring to patients and quality healthcare.” The United States government has established the Patient-Centered Outcomes Research Institute (PCORI) to investigate the relative effectiveness of various medical treatments. The intent may be to guide what sorts of therapies it will insure. A senior official from CMS (Centers for Medicare and Medicaid Services) recently acknowledged his department is indeed shifting to more outcome-oriented measures, as well as “a smaller set of measures that are meaningful to patients.” 1 With all these changes, how can physiatrists ensure they are receiving appropriate compensation for their services under this new pay-for-performance system?

Physical medicine and rehabilitation has a special interest in the comprehensive evaluation of persons with physical and/or cognitive functional limitations. Functional assessment was first defined by Lawton as “any systematic attempt to measure objectively the level at which a person is functioning in a variety of areas.”2 Physiatrists go beyond assessing patients’ biomedical health and seek to analyze and measure their functional health. Unlike other medical specialties in which a tumor is excised, a broken bone is repaired, or a heart valve is replaced, rehabilitation medicine’s benefits are not quickly or easily identified. Rehabilitation treatment may extend for days, months, or even years. How, then, do we identify and quantify the outcomes of our treatments? How do we accurately describe the benefits a patient has received from therapy? One solution is to use Precision Case Management (PCM), whereby biomedical information is combined with quantifiedpatient self-reports that serve as evidence of improved quality of daily living for patients, as is often the result of multidisciplinary care.

Precision Case Management and Self-Reported Evidence

PCM uniquely includes patients’ self-reports, or person-metrics, 3,4 which quantify how much chronic disease and disablement actually affect an individual’s quality of daily living. Having patients track their symptoms and quality of daily living creates authentic Self-Reported Evidence (S-RE) of their functioning in daily life. This is far better than presuming to “know what is best for the patient,” which is typically derived from biomedical data, objective tests, or simply the clinician’s interpretation of how the patient appears or responds. Using PCM, the patient and rehabilitation clinician work in partnership together to improve outcomes of care while reducing the use of unnecessary resources.

Essentially, PCM:

  1. Produces meaningful, valid, and useful information for guiding patient care.
  2. Demonstrates the effectiveness of rehabilitation treatment.
  3. Uses patient self-reports as evidence of the achievement of desired outcomes.

A recently published article supports this premise by suggesting that patient-reported outcomes offer the “potential to narrow the gap between the clinician’s and patient’s views of clinical reality and help tailor treatment to meet the patient’s preferences and needs.” 5

A variety of characteristics must be considered when treating each patient, including chronic or recurrent illnesses, personality, and environmental factors. The patient’s latent traits (favorable and unfavorable) may also play important roles. Because of the various ways they may be expressed, they are not easily identified; however, Rasch analysis may be used to help measure the effects latent traits have on the patient’s quality of daily living. Surveying and measuring a patient’s strengths and deficits also helps the physiatrist understand and appropriately treat the patient. The underlying question is whether the provided rehabilitation services are appropriate, necessary, and beneficial without causing concurrent worsening of the patient’s functional health. Both the patient and the clinician need qualitative and quantitative tracking of the illness, its treatment, and the effects latent traits have on the patient’s quality of daily living, such as:

  1. Experience with pain
  2. Participation in valued and necessary activities
  3. Biopsychosocial issues, including anxiety, depression, and spirituality
  4. Satisfaction and comfort with the levels of function achieved

Evaluations are required to be comprehensive.

A primary goal of medical rehabilitation is the comprehensive evaluation of a patient’s functional health. Measuring only at initiation of treatment is not sufficient—measurement should occur throughout and following treatment to demonstrate effectiveness. Successful achievement of goals often requires an assessment of the person’s functioning in a variety of life situations (e.g., in one’s home, work, school, and other community settings). Rehabilitation data and health assessment data are needed to identify treatments associated with more effective and efficient outcomes and to measure how much (and in what functional areas) patients are benefitting from care.

The Importance of Using Appropriate Measurement Techniques

Statistical theories commonly used for measurement are:

  1. Classic test theory (CTT): Based upon the total test, CTT considers a group of patients’ success rates on an item but does not inform about an individual patient’s success rate on an item.
  2. Item response theory (IRT): A popular method used to judge responses from a related set of observed items that represent a latent trait; however, IRT is not distribution-free and is descriptive, fitting the model to the data.
  3. Computer adaptive testing (CAT): An extension of IRT; it adapts. If an item is answered correctly, the computer follows with a more difficult question; if not, a less difficult question is presented.

While it is common for researchers to use various scales, tools and check-lists to describe samples of people with disablement, these tools have their drawbacks. For instance, they may not be uni-dimensional (describing a single functional concept) or equal-interval (linear rather than ordinal measurement), or their items may not be located in calibrated positions to form a hierarchy of difficulty. A measurement tool must be responsive to the attributes of an individual, not just groups of people, and should demonstrate changed responses of specific items over time in an individual with disablement. For more information, read: “What Is Measurement?” 6

Using Rasch Analysis to Measure the Effects of Latent Traits

Typically, a latent trait is a personal characteristic that is hidden or not obvious. As a result, it is often not acknowledged in the treatment plan. A person may have more than one latent trait. Some latent traits affect a person in a positive manner; others may be negative or troublesome. Negative latent traits can simulate a biomedical disorder, may vary in intensity, and often impose limitations on functional abilities.

Latent traits are attributes that cannot be measured with available tools such as rulers, thermometers, weight scales, etc. Georg Rasch was a Danish mathematician who developed a theory that enables reliable measurement of the biopsychosocial effects of latent traits, which are not customarily considered measurable. Rasch analysis (RA), or Rasch modeling, is a statistical method based on a mathematical theory for converting raw scores (such as the effects of latent traits) into true “probabilistic” measures that have the same validity as other measurement tools. The concept of having data fit the Rasch model rather than using a model that fits the data is foreign to most researchers who have been trained in classical statistics. Test validity is achieved when the data follow the requirements of the Rasch model. Most analysts are satisfied to derive a number that represents the sum of item values without exploring the behavior of the individual items that comprise the sum. Rasch modeling is unique in that it provides opportunities for uncovering information that should attract clinical attention on a case-by-case and item-by-item basis.

Rasch analysis maximizes the homogeneity of a trait and reduces redundancy without sacrificing information. Raw scores have unknown spacing between them. Rasch modeling builds estimates of true intervals of both item difficulty and subject ability. It calibrates item values and measures a subject’s abilities on a shared continuum that accounts for the latent trait. The improbability of a subject’s passing or failing is estimated item by item in terms of fit statistics. Fit statistics for items and subjects are identified as INFIT or OUTFIT, depending on whether the estimated value is near to or far from the expected value. 7,8,9,10

WINSTEPS Program for Rasch Analysis

WINSTEPS is one of several software programs available for Rasch modeling. Based on item calibration, it is an advanced method that provides objective, linear measurement. Rasch modeling transforms raw scores (ordinal ratings) and tests whether the data fit the model, as required of linear measurement. It is concerned with measuring individuals rather than a population’s distribution, and it transforms raw (ordinal) numbers with unknown distances between them into measures with equal-distance intervals (linear). All items within a measure have a single, shared dimension. Used appropriately, Rasch modeling accurately measures the effects of the latent trait components of functional health.

Analysts need to blend expertise in clinical, statistical, and Rasch measurement in order to successfully create useful and interesting measures of clinical phenomena. The biomedical model allows that symptoms and behaviors are expected to be proportional to pathophysiology. Although information from the history and physical examination, as well as laboratory data, is thorough, it may not sufficiently explain a patient’s observed symptoms and behaviors. When the relationships among pathophysiology, symptoms, and behaviors are clearly equivocal, they must be quantified and tracked with appropriate Rasch measures. Because biopsychosocial components often contribute to disablement, a patient’s functional health must be measured reliably. Reliable measurement should guide the clinician’s choice when selecting a course of treatment and judging the effectiveness of that treatment. Rasch-constructed measures constitute S-RE, which can help inform treatment decisions and judge the efficacy of treatment.

The LIFEwaresm System

UDSmr has incorporated Rasch analysis to create measures in the LIFEwaresm System, a robust data software application that works in conjunction with the Internet to produce self-reported evidence of a patient’s functional health in support of precision case management. We use quantifiable measurement, not just check-lists, of functional health status to track patients’ quality of daily living over time.

Designed for use in outpatient settings such as private practice offices or clinics, the LIFEwaresm System captures biopsychosocial information on the “whole person” for correct diagnosis, case management and treatment. It provides information on the effects of a wide variety of latent traits important to clinicians including: pain, depression, fatigue, behavior, role/social participation, cognition, memory, sleep, spiritual tendencies and satisfaction with treatment. Clinicians choose from a select group of impairment-specific patient assessments created from over 100 measures and scales. Repeat patient assessments throughout the course of treatment generate progress reports, which may justify adding to or adjusting the treatment plan.

The LIFEwaresm System avoids using raw-score ordinal ratings in favor of using linear measures to quantify functional health and to record the effects latent traits have on the patient. Quantified, descriptive data of outpatient treatment visits enables accurate tracking of the “ups” and “downs” of a patient’s functional health across several dimensions, as well as the patient’s responses to treatment over time—a relatively new development in medical management. The LIFEwaresm System not only identifies a patient’s functional status at the onset of treatment, but it also helps describe and quantify the functional outcomes achieved as a result of prescribed therapeutic interventions. Most importantly, it produces self-reported evidence of that improvement which can be shared with the patient, insurance provider or referral source.

Features of the LIFEwaresm System that have been in practice for over 25 years satisfy and even exceed the requirements of Medicare’s shift to Pay for Performance. The goal is precision case management (PCM). The following abbreviated case presentation provides examples.

A woman in her mid-50’s was a front seat passenger when the car veered off the road and struck a tree. Prior to the motor vehicle accident (MVA), X-rays had shown a transitional L5 vertebral body. Prior to the MVA, she enjoyed outdoor activities. After the MVA, she was started on physical therapy. Her gait remained markedly guarded, lumbar motion was limited, right ankle dorsiflexion and plantar flexion were mildly weak, and the right ankle reflex was reduced. For right sacroiliac joint tenderness, an injection was given. Associated with a post-traumatic stress disorder (PTSD), she experienced nightmares and flashbacks which cleared after a few months. However, insufficient sleep continued and remained chronic.

Pain medication was given only for the first few months and she was sustained on medication for tension and insomnia.

At 8 months post-injury she was referred for an Independent Medical Examination (IME). After examination the impression was: post-traumatic residua of sacroiliac joint dysfunction with piriformis muscle compression of the sciatic nerve causing discomfort and absent Achilles reflex. Uncoordinated ankle motions and symptoms had improved with physical therapy. Due to her ability to straight leg raise and hyperextend the spine, compression at the nerve root level was deemed unlikely. Chronic sleep insufficiency persisted despite evaluation and treatment.

The LIFEwaresm System was used to produce a comprehensive self-report of evidence covering the subject’s functional health and quality of daily living. Patient responses were rated on a zero to 100 scale. Only the measure of satisfaction with life in general and the pain scale had values below previously established standards. In the LIFEwaresm System, each item within a measure has an established expected value. The range of measures chosen for this case included: physical limitations, sleep level, body movement and control, driving, effort exerted, numerical pain intensity, verbal pain intensity, degrees of pain by types, mood characteristics, levels of distress, and satisfaction with function and results of treatment.

During the course of care, data were supplemented with the Spiritual Tendencies Inventory (STI) – *copyright C. Stephen Byrum, PhD and the Participation Profile of frequency and importance for twenty commonly experienced activities. The STI helps to identify whether an underlying personal characteristic is facilitating or else impeding improvement. The Participation Profile tracks the extent to which the person’s quality of life and sense of fulfillment are restored.

In this case the clinical course was monitored over 3 ½ years. Satisfaction with work participation improved from 80 to 100, satisfaction with homemaking remained at 90, and satisfaction with socialization improved from 70 to 100. However, sleep satisfaction declined from 70 to 60.

In summary, precision case management (PCM) includes use of self-reported evidence (S-RE) as a tool for monitoring, assessing and improving functional health (concurrent with attending to biomedical health). Areas for comprehensive monitoring include psychological/emotional/ spiritual functioning, satisfaction with treatment, societal and role integration, physical functioning including motor activities, pain, fatigue, cognitive functioning and memory.

International Classification of Functioning, Disability and Health

The LIFEwaresm System also incorporates the International Classification of Functioning, Disability and Health (ICF). 11 The ICF is a companion to the World Health Organization’s International Statistical Classification of Diseases and Related Problems (ICD), which forms the basis for coding and classifying medical conditions in the US and throughout the world. Whereas the ICD helps to classify medical conditions or diseases, the ICF classifies and codes human functioning—in particular, how one’s functional abilities are affected by the presence or absence of medical conditions and diseases. The ICF was endorsed for use in World Health Organization (WHO) Member States as the international standard for describing and measuring health and disability at both the individual level and the population level.

The ICF captures the interaction of body structure and function (impairments) with activities (limitations) and participation (restrictions) in the context of an individual’s unique environmental and personal factors and health conditions.

Nine components of the ICF indicate comprehensive measurement of functional health:

  1. Activity, Somatognosis, Mobility, Selfcare, Vision
    ICF codes: d1550, Acquiring skills; d2200, Carrying out multiple tasks; d4109, Changing basic body position, unspecified; d4259, Maintaining a body position, unspecified
  2. Pain, Fatigue, Energy
    ICF codes: b280, Sensation of pain; b28019, Pain in body part, unspecified; b4552, Fatiguability
  3. Mood, Affect, Sleep
    ICF codes: b1269, Temperament and personality functions, unspecified; b1529, Emotional functions, unspecified; d299, General tasks and demands, unspecified, b1349, Sleep functions, unspecified
  4. Social interaction, Interpersonal relationships
    ICF codes: d710, Basic interpersonal interactions; d720, Complex interpersonal interactions
  5. Role performance
    ICF codes: d740, Formal relationships; d750, Informal social relationships; d760, Family relationships; d699, Domestic life unspecified
  6. Cognition, Memory
    ICF codes: b117, intellectual functions, b144, memory function; k199, learning and applying knowledge, unspecified
  7. Behavior, Value tendencies
    ICF codes: d7202, Regulating behaviors within interactions; k7209, Complex interpersonal interaction, unspecified
  8. Satisfaction with comfort and function (achieved so far)
    ICF codes: b152, Emotional functions
  9. Participation
    ICF codes: d2208, Undertaking multiple tasks, unspecified

The Future

Going forward, it would not be surprising to see payments linked to patients’ functional improvement based on the ICF (“pay-for-performance”). After all, it makes more sense to base payment for services on results (functional improvement) rather than on the classification of the disease or medical condition (ICD) that caused the impairment. A dramatic change will occur when clinicians are reimbursed for the functional results they achieve from their treatment interventions (ICF) rather than being reimbursed for the diagnosis/medical condition the patient presents with (ICD).

Summary

As we function, so shall we live. The medical community must not just treat an injury or specific illness, but must consider other factors that may be affecting a patient’s functional health. The LIFEwaresm System provides valuable documentation of a patient’s functional health status and progress in an outpatient setting. This information can be generated for a physiatrist or an orthopedist during early recovery or even many years after injury. It also may be useful for a urologist, cardiologist, or another specialist who is consulted for other complaints that require treatment. It is always the job of the physiatrist to keep function foremost in the evaluation and management of patients. When functional health is not measured in outpatient settings, it is more difficult to acknowledge or manage.

Functional health outcomes demonstrate the true value physiatrists and multidisciplinary therapists bring to patients. Outpatient rehabilitation interventions must show measurable benefits which enhance quality of daily living for patients and their caregivers, thus justifying physiatrists and rehabilitation team members being paid for performance.

LIFEware is a trademark of Uniform Data System for Medical Rehabilitation, a division of UB Foundation Activities, Inc.

REFERENCES

  1. Lowes R. ‘Historic’ timeline set for basing Medicare pay on ‘value’. Medscape, Jan 26, 2015.
  2. Lawton MP. The functional assessment of elderly people. J Am Geriatr Soc. 1971; 19(6):465-81.
  3. Tesio L. Measuring behaviours and perceptions: Rasch analysis as a tool for rehabilitation research. J Rehabil Med. 2003; 35(3):105-115.
  4. Tesio L. Measurement in clinical vs. biological medicine: the Rasch model as a bridge on a widening gap. J Appl Meas. 2004; 5(4):362-366.
  5. Nelson EC, Eftimovska E, Lind C, Hager A, Wasson JH, Lindblad S. Patient reported outcome measures in practice. BMJ. 2015; 350:g7818.
  6. http://www.udsmr.org/Documents/What_Is_Measurement_2008.pdf
  7. Andiel C. Rasch analysis: a description of the model and related issues. Can J Rehabil. 1995; 9(1):17-25.
  8. Embretson SE. The new rules of measurement. Psychol Assess. 1996; 8(4):341-349.
  9. Wright BD, Stone MH. Best test design. Chicago, MESA Press, 1979.
  10. Granger CV, Carlin M, Linacre JM, Mead R, Niewczyk P, Stenner AJ, Tesio L. Rasch-derived latent trait measurement of outcomes: insightful use leads to precision case management and evidence-based practices in functional healthcare. J Appl Meas. 2010; 11(3):230-243.
  11. Granger CV, Brownscheidle CM, Carlin M, Graham JE, Malik C, Markello S, Niewczyk PM, Ottenbacher K, Tesio L. 2010. Functional Assessment. In: JH Stone, M Blouin, editors. International Encyclopedia of Rehabilitation. Available online: http://cirrie.buffalo.edu/encyclopedia/en/article/44/

Author Disclosures

Carl V. Granger, MD,
Nothing to Disclose

Lynne M. Adamczyk, RN, BSN, MBA
Nothing to Disclose

Related Articles