Print This Page

BACKGROUND

The American Geriatrics Society (AGS) strongly endorses efforts to ensure that every older person receives high-quality, patient centered health care. Pay for performance has the potential to improve care to this population. Current payment systems do not consider quality in determining reimbursement. Moreover, the incentives of the current reimbursement systems sometimes promote poor quality care. The present fee-for-service payment systems pay providers based on the number and complexity of services provided to patients without regard to quality, efficiency, or impact on health outcomes. Pay for performance has been proposed as one strategy designed to correct this deficiency.

However, AGS also recognizes that if the unique needs of this population are not appropriately taken into account, pay for performance could actually result in a reduction in quality, particularly to the most vulnerable and frail older adults. This Position Statement was developed to provide policymakers and health plan administrators assistance in moving payment toward a value based purchasing system in order to achieve greater quality in health care outcomes for all older adult patients, including the frail elderly.

POSITIONS

The following principles are essential to realizing the potential benefits of a pay for performance system for frail elderly.

1. A value-based purchasing system for the Medicare system must address the care of the large portion of Medicare beneficiaries who have multiple chronic conditions, are frail, of advanced age or require palliative care and not focus only on the care provided to the typical beneficiary. For these beneficiaries, measures should account for comorbidities and assess aspects of health that are common to multiple conditions (e.g., cognitive status, functional status, and pain.) Measures should be constructed so that providers are not penalized when they honor patients' preferences for care or their cultural or religious beliefs.

Rationale: The older adult population served by Medicare is extremely heterogeneous. Many people are healthy and functional, but up to 1/3 of the Medicare population are vulnerable, with multiple comoribidities and geriatric conditions (functional and cognitive impairment, falls, and frailty). In addition, some older adults will place different values on participation in self-management and adherence to medical recommendations, especially in the setting of multiple comorbities and health status vulnerability. Some older adults may put more emphasis on improved functioning and quality of life rather than traditional indicators of clinical care quality.

It is essential that a pay for performance program not unwittingly lead to a decrease in quality for vulnerable elders or those who may have different clinical care goals. Assessing and rewarding performance using indicators that have been developed for a commercially insured population and may not be relevant to vulnerable older adults has the potential to detract attention away from essential care and services. A pay for performance system needs to produce quality results that are meaningful and appropriate to the overall goals of clinical care for the patient population as a whole and for particularly vulnerable populations including the frail elderly. Failing to take this important policy concern into account could adversely affect access to primary and specialty care among those Medicare beneficiaries who might benefit the most from high-quality care.

Reference: Institute of Medicine. 2001. Crossing the quality chasm. A new health system for the 21st century. Washington, DC. National Academy Press.

Starfield B, Powe NR, Weiner JR, et al. Costs vs. Quality in different types of primary care settings. JAMA. 1994;272:1903-1908.

Roper WL, Winkenwerder W, Hackbarth GM, Krakauer H. Effectiveness in health care. An initiative to evaluate and improve medical practice. NEJM. 1988:319:1197-202.

Medpac. Report to Congress - Chapter 4: Strategies to improve care: Pay for performance and information technology. March 2005.

2. Payment reform is needed to drive practice change. Effective pay for performance systems support and stimulate the structural capabilities necessary for the provision, documentation, measurement and continuous improvement of high quality care. The need for different provider settings to care for unique patient populations must also be considered when designing and implementing payment based upon performance. For example, larger practices may have more resources to implement quality care processes. Such resources include providing patient education materials for patients and their caregivers, language translation services, and other outreach activities to facilitate good patient care. A smaller practice may have fewer resources to measure and report quality, yet be an essential care provider, such as a rural care provider or home bound elderly care team. Certain practices may be focused on a population subset for which there is no performance measures at this time. Payment reform should lead to improved practice design without eliminating essential care providers.

Rationale: Providers who use innovative approaches that improve quality of care find that most current payment systems do not provide them with the resources needed to sustain these activities. Today's fee for service payment system does not provide payment for services such as health education or development of the infrastructure to measure and report quality. The result is that providers are unable to invest in activities that have great potential to improve quality and avoid unnecessary medical costs. An example of such activities includes health information technology (HIT) including electronic health records and ePrescribing. Providers typically cite the cost of these HIT systems and lack of a clear return on investment as barriers to their implementation. In addition to incentives of higher reimbursement for provision of better quality care, funds should be made available through direct grants and low interest loans for the infrastructure responsible for implementation of an HIT system within a practice. HIT systems are not only used for the management of clinical records, but are critical elements in data collection, clinician feedback, quality improvement, and population tracking. Each of these is essential to practice improvement as the final desired outcome of pay for performance programs.

Reference: McClellan M. Letter to the Honorable William M. Thomas, Chairman Committee on Ways and Means. June 24, 2005.

Bates DW, Gawande AA. Improving safety with information technology. NEJM. 2003. 348:2526-2534.

Miller R, Sim I. Physicians' use of electronic medical records: Barriers and solutions. Health Affairs. 2004. 23;2:116-26. Medpac. Report to Congress - Chapter 4: Strategies to improve care: Pay for performance and information technology. March 2005.

3. Structure, process and clinical outcomes measures used must be valid and relevant for the unique care needs of frail or vulnerable older adults. These measures should be evidence-based, clinically relevant, have clear association with improved outcomes of care, and be applicable to all patients whose care they assess.

Technical specifications for numerators and denominators of measures should be constructed so that measures are not applied to special populations where evidence of linkage of performance of care processes to improved outcomes is lacking. These populations include persons of very advanced age, those with multiple comorbidities, limited life expectancy, or moderate to severe dementia.

More importantly, the AGS believes that specific measures are needed to assess the quality of care of persons over age 75, those who are vulnerable and/or frail and who are receiving palliative care near the end of life. Clinical performance measures relevant to the Medicare population are of three types:

1. Structure measures - used to recognize systems of care associated with improved health outcomes. Multi-disciplinary teams for care, capacity for patient education in self-care management, disease registries, electronic health records and systems to support the use of inter-visit interval patient contacts and monitoring are important aspects of the chronic care model. Important processes of care are difficult to deliver absent such structure(s). Initial reward systems should recognize investment in such delivery structure.
   
   
2. Process measures - used to determine whether care that is known to be effective is provided. The AGS believes that specific measures are needed to assess the quality of care of persons over age 75, those who are frail or vulnerable and those who are receiving palliative care near the end of life. The Assessing the Care of the Vulnerable Elders (ACOVE) measures were developed by the RAND Corporation in response to the needs of persons at risk for frailty or persons aged 75 and older. They have been tested in cohorts of community-dwelling vulnerable older adults. Adoption of some or all of these indicators in addition to disease and prevention based measures appropriate for the younger population will greatly enhance attaining the goal that quality of care can be measured for all Medicare beneficiaries and that the providers who care for these patients have measures of accountability.
   
   
3. Clinical Outcome measures - must be appropriate to the usual health care needs and goals of the patient population in which they are used. Many of these measures are disease-specific, originally developed in commercially insured populations, and do not account for other comorbidities. If not previously tested in the heterogeneous Medicare population, such measures may be particularly problematic. In frail or vulnerable older persons, the goals of treatment for chronic disease are more variable than in younger adults, and the linkage between process of care delivered and clinical measures is often imperfect.

Rationale: High quality geriatric care requires providing services to a heterogeneous population with varied health care service needs and goals of care, including many with multiple chronic illnesses and geriatric conditions. Fundamental to the definition of quality is that the standard of care being measured is applicable to the individual to whom it is applied. Many available clinical performance measures have been developed in the middle-aged commercial population. Some of these measures have been tested in older adult populations, but some have not, so the applicability of untested measures to older adult, heterogeneous populations is not known. The complexity of delivering high quality care to this population can only be guided by rigorous scientific testing of performance indicator measures. Such testing can allow us to adapt measures to the specific needs of vulnerable patient populations. These measures need to be developed and validated by individuals with expertise in the care of frail elders. Measurement development and testing is a dynamic process so even in the setting of value-based purchasing there should be an established process for continual evolution of performance indicator measures. Key stakeholders should remain involved, including the American Geriatrics Society.

Reference: Wenger NS, Shekelle, et al. Assessing care of vulnerable elders: ACOVE project review. Annals of Internal Medicine. 2001. 135;8:642-646.

AGS Ethics Committee. Measuring quality of care for nursing home residents - considering unintended consequence. Annals of Long-term Care. 2005. 13(8):40-44.

Boyd CM, Darer J, Boult C, Fried LP, Boult L, Wu AW. Clinical Practice Guidelines and Quality of Care for Older Patients with Multiple Comorbid Diseases: Implications for Pay for Performance. JAMA 2005;294:716-724.

E.H. Wagner, B.T. Austin and M. Von Korff, "Organizing care for patients with chronic illness," Milbank Quarterly 74 (1996): 511-544 abstract link on pubmed: Click Here

4. A pay for performance system must provide positive reinforcement for quality performance and improvement and not promote the avoidance of patients for whom providing high quality care will be more challenging. When applied to individual providers, these measures should be constructed so that both achievement of target thresholds (excellence) and progress toward achievement of targets (improvement) can be positively reinforced. In addition, the collection of data that are used to evaluate performance should not be burdensome to providers and should accurately reflect the performance of care processes on an individual patient basis. Failure to consider these factors may unduly penalize providers who practice in small groups or care for special populations.

Rationale: Linking a portion of payments to valid measures of quality and effective use of resources will result in providers having direct incentives and financial support to implement the innovative approaches that result in improved outcomes. All monies that are set aside for pay for performance should be distributed to providers achieving the quality criteria. Savings from improved care are likely to accrue to Medicare funds that are not part of the physician resource pool. Performance based funds should not be used in a withhold approach or a Medicare Part B budget neutral manner. To the fullest extent possible, norms should be established for like populations and risk adjustment should occur. While it is recognized that value based purchasing should not be delayed while such adjustments are refined, such an approach will minimize the risk of providers avoiding patients (e.g., the frail and medically complex) who present the greatest challenge in meeting quality indicators used for pay for performance.

Reference: Roland M. Linking physicians' pay to the quality of care - A major experiment in the United Kingdom. NEJM. 2004. 351(14):1448-1454.

Agency for Healthcare Research and Quality, Department of Health and Human Services. 2004. Strategies to support quality-based purchasing: A review of the evidence. Pub. No. 04-P024 technical review 10. Rockville, MD: AHRQ. July.

CREDITS

American Geriatrics Society and approved by the AGS Board of Directors in January 2006. Written by the AGS Health Care Systems Committee and Quality Ad Hoc Workgroup, with special thanks to Drs. Richard Stefanacci, Eric Coleman, Cheryl Phillips, Peter Hollmann, Meghan Gerety, and Caroline Blaum. AGS, The Empire State Building, 350 Fifth Avenue, Suite 801 New York, NY 10118.