Geographic Adjustment in Medicare Payment: Phase II: Implications for Access, Quality, and Efficiency (2012)

4

Programs and Policies to Improve Access
and Quality of Care for Beneficiaries

INTRODUCTION

This chapter examines available evidence on the extent to which various programs and policies have influenced the ability of U.S. communities to attain adequate access to services appropriate to meet their health care needs. Previous chapters have established that access to high-quality health care services, including primary care services delivered by physicians and other practitioners, is not uniform across the United States. The health care workforce varies in size relative to population, and it varies in composition both across and within metropolitan and nonmetropolitan areas. In spite of this variability, broadly speaking, access to health services and Medicare beneficiary satisfaction with access are generally adequate and comparable in rural and urban areas.

The differences in health workforce size, distribution, and composition have been recognized for many years, and programs have been developed, mostly within the Public Health Service, to address them. The evidence reviewed by the committee suggests that geographic access to health care services has been improving, most likely as a result of market forces as well as various workforce policies, but that the distribution of practitioners continues to be a concern (Ricketts and Randolph, 2007, 2008; Rosenthal et al., 2005).

At the same time, factors that affect practitioner compensation, including payment policies of the Medicare program, may also have had an effect on the health care workforce’s ability to provide acceptable access in different geographic areas. Among the newest Medicare payment policies are those that move practitioner compensation from traditional fee-for-service models to bundled payments and other incentives for care teams to coordinate care across organizations and settings. These policies are encouraging new delivery models that are intended to improve efficiency and provide a better quality of patient experience, but their likely influence on the workforce supply, distribution, and training programs is not yet clear.

In the statement of task for Phase II, the committee was asked to evaluate and consider the effect of the geographic adjustment factors on the level and distribution of the health care

workforce, including recruitment and retention, mobility between urban and rural areas, and the ability of hospitals and other facilities to maintain an adequate and skilled workforce in order to maintain access for beneficiaries. The committee was also asked to consider the effect of the adjustment factors on population health, quality of care, and the ability of providers to furnish efficient, high-value care.

Historically, policies and programs supported by the Centers for Medicare & Medicaid Services (CMS) and the Health Resources and Services Administration (HRSA) have sought to improve access to care in many different ways. In general, however, the policies have influenced the workforce directly, through training programs and payment policies to increase supply, or they have been targeted to maintain access through influencing the geographic distribution of facilities and health professionals. Medicare payment policies have also been tied to quality improvement for several years, originally for hospitals and increasingly also for ambulatory care and primary care providers, primarily physicians. The committee did not find sufficient evidence about the effect of payment policies on population health and high-value care to be able to include in its review, but it did discuss value-based purchasing and the workforce implications of new models of care that focus on care coordination.

Given the breadth of the committee’s charge, the committee chose to do a targeted review that focused on Medicare payment policies to address access, quality of care, and workforce supply and distribution and that also have a geographic component, such as a comparison of metropolitan and nonmetropolitan areas. This chapter begins with a review of Medicare policies and programs intended to promote beneficiaries’ access to hospital and primary care services provided by a variety of health professionals, and then it reviews policies intended to promote quality of care. The chapter then reviews workforce programs intended to improve the geographic distribution of practitioners through recruitment and retention efforts, focusing on program evaluations and other evidence that the programs are successful in improving access, especially in Health Professional Shortage Areas (HPSAs). The chapter then discusses the many gaps in the evidence it reviewed and the need for a coordinated approach to collecting workforce data, designing programs, and setting national workforce targets and goals. The chapter closes with the committee’s findings related to access, quality of care, and workforce programs and policies.

MEDICARE PAYMENT POLICIES INTENDED TO MAINTAIN ACCESS TO HOSPITAL CARE

Given the committee’s focus on the impact of geographically based payment adjustments on access in medically underserved areas, hospitals that are important or sole sources of hospital care for Medicare beneficiaries were of particular concern. Medicare’s payment policies that are intended to preserve access to hospital care in geographically isolated areas focus on five types of hospitals: critical access hospitals, sole community hospitals, Medicare-dependent hospitals, low-volume hospitals, and rural referral centers (see Table 4-1). However, the policies that apply to these hospitals tend to be inconsistent, and there is no mechanism for ensuring that the policies serve their stated purpose.

Nearly 1,300 hospitals have been designated as critical access hospitals, based on their size and the lack of another hospital within a specified distance.¹ The critical access program is

______________

¹ Medicare Payment Advisory Commission. Payment Basics: Critical Access Hospitals Payment System, revised October 2011. Available at http://www.medpac.gov/documents/MedPAC_Payment_Basics_11_CAH.pdf. Critical access hospitals are limited to 25 acute care beds and must be at least 35 miles by primary road or 15 miles by secondary road from the nearest hospital; until 2006, hospitals also could qualify as critical access hospitals if they were designed a “necessary provider” by their state.

TABLE 4-1 Access to Hospital Care in Geographically Isolated Areas

NOTE: CMI = case-mix index; FY = fiscal year; IPPS = Inpatient Prospective Payment System.

designed to maintain access to emergency care and limited hospital inpatient services in isolated rural communities that are unable to support a full-service hospital. Critical access hospitals are paid based on their current Medicare allowable costs; because the committee’s recommendations apply only to those hospitals that are paid under Medicare’s inpatient prospective payment system for hospital services, critical access hospitals are not affected by the committee’s recommendations on the hospital wage index.

The special protections afforded to about 400 sole community hospitals are intended to support their unique role in providing access to inpatient hospital care to the residents of a geographic area. Medicare’s payment to sole community hospitals is based on the higher of the applicable standard Inpatient Prospective Payment System (IPPS) rate or a hospital-specific rate derived from the hospital’s own historical costs updated for inflation (using the Medicare hospital market basket index) and adjusted for case-mix changes. Sole community hospitals

currently paid at their hospital-specific rate generally are not affected by the committee’s recommendations; for sole community hospitals that currently are paid their standard IPPS rate, the committee’s recommendations could reduce or increase their payments, but in any case they could not be paid less than their hospital-specific rate.

The approximately 160 Medicare-dependent hospitals are afforded similar but more limited protection, to recognize their presumed inability to cover any difference between their Medicare costs and the standard IPPS rates. If the IPPS rate is lower than the hospital-specific rate (derived as for sole community hospitals, as described above), a Medicare-dependent hospital is paid 75 percent of the difference between the IPPS rate and its higher hospital-specific rate. Unlike critical access hospitals and sole community hospitals, Medicare-dependent hospitals are not designated based on the lack of alternative sources of care available to Medicare beneficiaries, but only on their small size and high proportion of Medicare patients. As a result, this program is not as well targeted to preserving access to care in geographically isolated areas.

The rural referral hospital designation was established when it was determined that large rural hospitals that served as tertiary referral centers were disadvantaged by IPPS rate-setting policies that did not adequately account for their more complex patient population and infrastructure costs. Rural hospitals received a lower standard payment rate than urban hospitals; the patient classification system used to determine payment did not adequately account for differences in severity; and the hospital wage index reflected relative wage levels with no occupational mix adjustment. The current prospective payment system, however, applies the same payment rate to hospitals located in rural areas and urban areas with less than 1 million population, has improved its ability to account for differences in patient severity, and provides for a limited occupational mix adjustment to the hospital wage index (which the committee’s recommendations would further improve). The approximately 130 rural referral centers also benefit from less stringent geographic reclassification standards (which the committee’s recommendations would eliminate, in any case) and they also may receive higher disproportionate share payments than small urban and most other rural hospitals. The changes in Medicare payment over time would seem to weaken the rationale for the establishment of this category of hospitals for purposes of payment.

The Medicare Modernization Act of 2003 established a payment enhancement for low- volume hospitals located more than 25 road miles from another hospital and having fewer than 800 total discharges. On the grounds that they cannot achieve the economies of scale and scope of larger hospitals and therefore tend to have higher costs per discharge, they can receive up to a 25 percent increase in their IPPS payments based upon volume.

While these payment policies may be intended to preserve access to needed hospital care, they could potentially be better targeted to efficiently and effectively meet this objective. For example:

•   The standards used to identify geographically isolated hospitals vary from one provision to another. The definition of road miles has been standardized, but the number of road miles differs from 35 miles under the sole community hospital policy to 25 miles under the permanent low-volume adjustment (and 15 miles under the temporary policy). The criteria for sole community hospitals consider the reliance of Medicare beneficiaries on the hospital while the low-volume adjustment criteria do not. Some hospitals receiving the low-volume adjustment may not be needed to preserve Medicare beneficiary access to care.

•   There is no periodic review to confirm whether hospitals designated as sole community hospitals and rural referral centers continue to meet the qualifying criteria for the higher payments. Moreover, rural referral centers located in a county that is redesignated from rural to urban by the Office of Management and Budget are permitted to retain their rural referral designation regardless of whether there are other hospitals in close proximity.

•   In general, several of these policies may not be fulfilling their intended purpose. The hospital-specific rate for a sole community hospital is based on the highest cost per discharge from 1982, 1987, 1996, or 2006 updated for inflation in input prices. Using a hospital’s costs in a more recent year to establish the hospital-specific rate would result in a better match between the hospital’s current cost structure and Medicare’s payment amount. As described above, the rationale for rural referral centers is substantially weaker than it was when the IPPS was first implemented. With regard to the low-volume adjustment, the temporary discharge criterion based on Medicare (rather than total discharges) eliminates the empirical underpinning for the adjustment and disadvantages hospitals with high Medicare utilization relative to hospitals of comparable size with low Medicare utilization rates.

•   There is some redundancy among the policies. Sole community hospitals (and Medicare-dependent hospitals) are eligible to receive a low-volume adjustment on their hospital- specific rates as well as their IPPS rate. Because the former should already reflect the cost effect of providing a low volume of services, a low-volume adjustment to the hospital- specific rate is unnecessary.

The committee is concerned that these considerations diminish the effectiveness of Medicare policies in ensuring access of Medicare beneficiaries in different areas to appropriate care.

PROGRAMS THAT ENCOURAGE CLINICAL PRACTICE IN UNDERSERVED AREAS

Community Health Centers

Since 1965, the U.S. government has funded community health centers (CHCs) to provide primary care services to underserved populations in metropolitan and nonmetropolitan areas, including low-income and uninsured populations. More than 1,100 CHCs operate more than 8,100 delivery sites that care for a total of 19.5 million individuals in every state, the District of Columbia, Puerto Rico, the U.S. Virgin Islands, and the Pacific basin (HRSA, 2012d).

As of 2006, primary care physicians accounted for 89 percent of all physicians working in CHCs (Rosenblatt et al., 2006). CHCs also rely on advanced practice nurses and physician assistants (PAs) for staffing. In 2010, CHCs employed more than 131,000 staff nationally, including 9,600 physicians; 6,400 nurse practitioners (NPs), physician assistants, and certified nurse midwives; 11,400 nurses; 9,500 dental staff; 4,200 behavioral health staff; and more than 12,000 case managers, health educators, outreach workers, and transportation staff (UDS, 2012; The White House, 2012).

CHCs substantially rely on the incentives for health professionals to work in these settings provided by loan forgiveness programs and J-1 visa waivers, which will be discussed later in this chapter. In 2006, CHCs had large numbers of unfilled positions, notably for family physicians, obstetricians/gynecologists, and psychiatrists (Rosenblatt et al., 2006).

A substantial amount of research has been conducted on quality of care in CHCs, but there

are no direct comparisons of CHC quality relative to comparison groups of private practices. The limited evidence that exists indicates that quality of chronic disease care is comparable to other settings (e.g., Hicks et al., 2006; Landon et al., 2006). However, similar to general practices, the quality of care in CHCs varies across settings, with some outcomes comparable to national benchmarks, while others do not always meet quality-of-care standards and guidelines (Chin et al., 2000). Studies of CHC interventions to improve clinical workflow and patient education for diabetes patients have resulted in some improvements in clinical outcomes (Chin et al., 2007).

Area Health Education Centers

Area Health Education Centers (AHECs) recruit, train, and retain health professionals to work with underserved populations by developing partnerships among medical, nursing, and allied health schools to help meet local health care needs (HRSA, 2012c). Fifty-nine AHEC programs and more than 245 rural and urban centers operate in 48 states, 2 territories, and the District of Columbia.

The federal government has supported AHECs since 1971 and administers the program through HRSA. The AHEC program grantees are medical and nursing schools who contract with CHCs to provide clinical rotations and training opportunities for health professionals who seek experience providing clinical care, health education, and preventive services for underserved communities. They also provide infrastructure and a combination of support services, including medical library resources, continuing education courses, and telecommunications technology linking clinical practices in rural areas with personnel in an academic medical center or nursing school.

The Office of Inspector General (OIG) (1995) of the Department of Health and Human Services (HHS) conducted a short-term management and program evaluation of four AHECs in Arkansas, Florida, and Texas. The key results, somewhat dated given that the evaluation was performed over 15 years ago, were that the AHECs were responding to clinical needs of practitioners in rural areas but that there was a need for greater emphasis on educating practitioners about innovations in health care delivery, such as clinical practice guidelines, and opportunities for use of telecommunications technology already available at the time were not being sufficiently used.

To the committee’s knowledge, there has been no systematic evaluation of AHECs, although there are qualitative reviews of specific aspects of particular programs (see e.g., Rooks et al., 2001) on a primary care preceptorship for first-year medical students coordinated by an AHEC based at the University of Florida.

Incentive Payment Program for Primary Care Services in HPSAs

A total of almost 60 million Americans, or about one in five, live in geographic areas that are designated as primary care shortage areas (HRSA, 2012a). Since 1987, Section 1833(m) of the Social Security Act has provided bonus payments for all services for physicians in locations designated as primary medical care HPSAs under Section 332 (a)(1)(A) of the Public Health Service Act. In 1991, the original 5 percent bonus was increased to 10 percent. This bonus is applied to cover Medicare Part B services provided in designated geographic HPSAs. For claims with dates of service on or after July 1, 2004, psychiatrists providing services in mental health HPSAs are also eligible to receive bonus payments.

From 2011 through 2015, participating Medicare Part B primary care practitioners will receive an additional bonus payment equal to 10 percent of the amount paid for primary care services.² Eligible primary care practitioners include physicians (family medicine, internal medicine, geriatric medicine, or pediatric medicine), NPs, clinical nurse specialists, and PAs. In addition, the primary care services they provide must account for at least 60 percent of allowed charges under the Physician Fee Schedule in order to receive the 10 percent bonus payment (CMS, 2011b). The threshold for the bonus excludes some clinicians who see fewer Medicare patients and provide services to fewer beneficiaries than CMS requires.

There has been very little empirical research on HPSA bonuses. One study by Chan et al. (2004) used Medicare Part B claims data for 1998 to assess amounts paid under the 1991 10 percent bonus program and reached two important conclusions. First, amounts paid were small, which could be expected to limit incentive effects of the program. Second, many claims that could have been subject to the bonus had no bonus payments, and many bonus payments that should not have been paid were paid. Shugarman and Farley’s (2003) study contributed additional evidence on the small size of the 1991 bonus program, documenting that bonus payments constituted about 1 percent of total Part B payments for services in rural, under-served areas. Given its small size, important effects on HPSA practitioner workforce could not be expected, and the same may be true of the Affordable Care Act bonus program. Given that the more recent bonus payments have only been available for about a year, it is too early to evaluate their effects on the primary care workforce. However, any temporary bonus can be expected to have a much lower effect on a clinician’s location choice than a permanent bonus, because of the length of time often involved in making location decisions.

HPSA Surgical Incentive Payment (HSIP)

The Affordable Care Act of 2010, Section 5501 (b)(4) provides bonus payments for general surgeons in HPSAs. Effective 2011 through 2015, physicians serving in designated HPSAs will receive an additional 10 percent bonus for major surgical procedures within a 10- or 90-day global period (CMS, 2011b). This additional payment, referred to as the HPSA Surgical Incentive Payment (HSIP) will be combined with the original HPSA payment and will be paid on a quarterly basis.

HSIP is intended to provide incentives for general surgeons in medically underserved areas, as they make it possible for many of the nation’s smallest and most remote hospitals to provide services (Hagopian et al., 2003). Many rural towns can lose non-hospital-employed physicians as well as those employed by hospitals after their hospitals close (Hart et al., 1994). Similar to the Affordable Care Act primary care bonus program, studies of the effects of the Affordable Care Act provision on the general surgery workforce are not yet available.

Methodological Questions About Provider Distribution

Current official measures of the distribution of physician personnel and by inference, geographic access to physicians’ services, such as HPSAs, are based on county boundaries. County

______________

² The bonus provisions were described in the Affordable Care Act of 2010. See http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/HPSAPSAPhysicianBonuses/index.html?redirect=/hpsapsaphysicianbonuses.

boundaries were established historically and are at best imprecisely related to medical market areas, because beneficiaries routinely cross county and other boundaries to obtain care.

Using data from 23 states with relatively low physician-to-population ratios compared to the other 27 states, and assuming that people seek care from nearby primary care physicians, Rosenthal et al. (2005) found that few persons would have relied on primary care physicians with caseloads in excess of federal guidelines for HPSAs in 1999 (i.e., caseloads exceeding 3,500 patients per physician). In particular, only 11 percent of individuals residing in rural counties adjacent to metropolitan areas would have obtained care from primary care physicians with caseloads exceeding this threshold.

For residents of rural counties not adjacent to metropolitan areas, Rosenthal et al. (2005) estimated that 7 percent saw primary care physicians with caseloads above the threshold represented in the federal HPSA guidelines. The authors took the position that many individuals in counties adjacent to metropolitan areas obtain care from providers in the metropolitan areas and hence should not be attributed to providers in their counties of residence. For this reason, the caseload burden in such counties is actually lower than their calculations imply.

Thus, using an alternative assumption about individuals’ travel to care, meaning they travel further than was assumed in the baseline calculations, reduces variation in caseloads by area, which implies less of a geographic maldistribution of providers. In other words, because so many individuals are willing to travel to see practitioners in both urban and rural locations, it is very difficult to estimate whether or not there is a maldistribution of clinicians on the basis of provider location alone.

Other results in the Rosenthal et al. (2005) study also raise issues about the strength of the relationship between provider location and availability of care. For one, residents of very rural counties had to travel 5 miles on average to the nearest general or family practitioner. The nearest specialist in internal medicine was slightly over three times further away, and physicians in specialties with fewer practitioners (e.g., ophthalmology and neurology) were even further away. Primary care physicians locate their practices in communities of all sizes, so a close link would be expected between primary care availability, especially general and family practitioners, and access to primary care services. By contrast, the geographic locations of physicians in smaller fields such as subspecialty care tend to be limited to communities with larger population sizes.

Availability of care depends on many factors, including work hours—total and scheduled hours, staff size and variety of personnel, willingness to accept new patients, including Medicare beneficiaries, and many others. As Rosenthal et al. (2005) pointed out, there are some limitations to their study. First, as they acknowledge, the American Medical Association Masterfile, which the authors used for data on the location and specialty of physicians, is a headcount. Second, they made assumptions about travel patterns for care but did not observe actual patterns. Any measure of central tendency, such as mean distance between a place of residence to the nearest physician, obscures important variation. For example, although it is important to know that a mean travel distance is 5 miles, travel is going to be much longer for those who live in isolated rural communities. Finally, workforce policies of the U.S. government may reduce differences in ratios of health professional personnel relative to population. By excluding federal physicians and other federal clinicians from their analysis, Rosenthal et al. (2005) may actually have overstated geographic variation in these ratios.

The Rosenthal et al. (2005) study makes an important contribution in raising questions about the degree of geographic maldistribution of physicians and in emphasizing the limitations of using counties as units for calculating shortages. As a matter of policy and practice, however,

additional studies will be needed to fully understand the range of influences on individual willingness to travel to see providers, and policy makers will need to conduct further studies and consider setting national distribution targets for the workforce in order to determine the extent of shortages.

CURRENT PROGRAMS TO IMPROVE QUALITY

As discussed in Chapter 3, HHS uses a core set of evidence-based quality measures to make standardized comparisons of the quality of care for different groups and regions (AHRQ, 2010). Quality measures include process measures, which are derived from administrative or claims and clinical data and measure the delivery of care; and outcome measures, which are often disease specific and include clinical outcomes as well as patient experience and satisfaction with the care team and the care setting. Hospital reporting on quality measures has been under way for several years, while primary care quality reporting is in earlier stages. However, several policies tie payment to performance on quality measures.

Hospital Inpatient Quality Reporting Program

For the past 6 years, CMS has been administering the Hospital Inpatient Quality Reporting Program,³ which is designed to incentivize IPPS hospitals, through Medicare Part A payments, to report their quality of care measures to CMS (CMS, 2009). In 2010, participating hospitals were required to report 42 quality measures, including 30-day mortality and 30-day risk- standardized readmissions on three specific medical conditions,⁴ patient safety indicators and hospital-acquired conditions, and patient satisfaction data (CMS, 2011c).⁵

Hospital patient satisfaction data are collected from each hospital by administering the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey. Developed by CMS and AHRQ, the HCAHPS is a national standardized survey that asks discharged patients 27 questions about their recent hospital stay (HCAHPS, 2012). Since 2008, each hospital has been responsible for administering the survey to a random sample of adult patients (18 years of age and older) between 48 hours and 6 weeks after discharge.

Participating hospitals that fail to report the 47 quality measures and the HCAHPS patient satisfaction data receive a 2 percent reduction in their annual market-based update in Medicare payment. In FY 2011, only 47 IPPS hospitals (or less than 5 percent of all IPPS hospitals) chose not to participate in the quality reporting program and therefore received a 2 percent reduction in their annual market-based update in Medicare payment (CMS, 2011d).

Hospital Readmissions Reduction Program

New Medicare payment policies are targeting payments to help reduce overall Medicare spending while maintaining or improving the quality of care. Based on evidence from pilot programs, demonstration projects, and expert consensus, these policies reflect a combina-

______________

³ As mandated by mandated by Section 501(b) of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003.

⁴ The three conditions are heart attacks, heart failure, and pneumonia—the three most common medical conditions of hospital-admitted Medicare beneficiaries.

⁵ Starting in 2013, each quality measure specified by the Secretary of HHS must be endorsed by a contracted consensus entity (currently the National Quality Forum) (P.L. 111-148).

tion of financial incentives for desired performance and penalties for poor performance or nonparticipation.

The first efforts at quality reporting and monitoring began with hospital services, which represent a relatively large share of Medicare outlays. In 2008, spending on hospital inpatient care came to $129.1 billion, or 29 percent of total Medicare payments that year. Despite such high spending, the quality of care was deemed not adequate for beneficiaries who had multiple chronic conditions and see multiple practitioners. Readmission rates varied substantially across the country, even after controlling for the severity of illness (Jencks et al., 2009). For example, Illinois, Louisiana, and New Jersey readmission rates approach 22 percent, while Idaho, Oregon, and Utah were between 13 and 16 percent (CRS, 2010).

Between 2003 and 2004, 19 percent of Medicare beneficiaries who were admitted to a hospital were readmitted within 30 days (Hansen et al., 2011; Jencks et al., 2009). The total cost of all rehospitalizations within 30 days has been estimated at $44 billion (Jencks, 2010), and MedPAC (2007) has estimated that 75 percent of these hospitalizations may be avoidable; preventing them could save Medicare $12 billion a year. Factors contributing to preventable hospital readmissions include poor coordination between different care settings (e.g., hospitals, skilled nursing facilities, outpatient clinics), such as a lack of follow-up appointments after discharge (Jencks et al., 2009), as well as lack of assistance and support for frail beneficiaries at home as they recover.

As part of a major national policy initiative, CMS made 30-day readmission rates publicly available on its Hospital Compare website in 2010 to try to bring increased attention to this measure of quality of care, which reflects not only the quality of inpatient care but the ability of the care system to coordinate postacute care as patients transition across settings, such as from hospital to nursing home to home.

Readmission rates have been found to vary by hospital and by geographic area, even for the same level of severity for the same disease. This variation suggests that some readmissions could be prevented if there were better care management, particularly at discharge as beneficiaries transition to other care settings, such as their homes, skilled nursing facilities, or other postacute care (CRS, 2010). Beginning in 2013, CMS will reduce Medicare Part A payments for hospitals with higher than expected risk-adjusted readmission rates for three conditions: heart attack, heart failure, and pneumonia (CMS, 2011h). Medicare payments will be reduced by an adjustment factor based on the ratio of aggregate payments for excess readmissions (determined as a function of spending) to aggregate payments for all discharges.

Over time, Medicare plans to expand this program to include other common diagnoses, which will be good for beneficiaries and for Medicare. However, there are no economic incentives for hospitals to reduce Medicare readmissions, and the financial penalties may not be strong enough to make a business case for improving quality. It remains to be seen whether other models such as bundled payments and “single-episode prices” may be more effective in improving efficiency and patient experience while lowering costs (Berenson et al., 2012).

Of further concern, African American and Hispanic beneficiaries are more likely to be rehospitalized for preventable conditions than are white beneficiaries. An analysis of hospital discharge data from 10 states found that African American and Hispanic patients were at greater risk of rehospitalization, even after adjusting for patient differences in health care needs, socioeconomic status, insurance coverage, and the availability of primary care (Gaskin and Hoffman, 2000). In a comparison of predictors for rehospitalization after coronary artery bypass surgery, African Americans had higher rates of readmission (Hannan et al., 2003).

The specific reasons for readmissions are likely to vary by individual and care setting, but cultural differences in care seeking as well as racial and ethnic differences in how clinicians and health care organizations care for patients may have an influence on the readmission rates in different areas of the country. CMS has drawn attention to the problem of hospital readmissions by including a risk-adjusted 30-day readmission rate for heart failure patients as one measure of the quality of care (CRS, 2010). However, the rates are not always reported by race and ethnicity, and documenting and addressing potential disparities will require further attention.

Payment Policies and Quality Measures for Physicians and Other Providers

Since 2006, CMS has been collecting quality data measures from physicians and other eligible practitioners⁶ paid under the Physician Fee Schedule under the Physician Quality Reporting System (PQRS). Similar to the Hospital Quality Reporting Program, the PQRS provides financial incentives to eligible Medicare practitioners who successfully submit the required quality measures to CMS. In CY 2011, eligible practitioners who satisfactorily reported data on 175 individual quality measures and 13 quality measure groups received a 1 percent incentive payment of their total estimated allowed charges (CMS, 2011j). For 2012 through 2014, eligible professionals may earn an incentive payment of 0.5 percent of their total estimated allowed charges (CMS, 2011j).

The PQRS began as a voluntary program, but beginning in 2012, practitioners who do not report quality data will have their annual Medicare payment update reduced by 2.5 percent (CMS, 2011j). Beginning in 2014, clinical quality reporting will be required. Also beginning in 2012, the secretary of HHS will be required to provide clinician feedback reports that compare each clinician’s resource use to the resource use of other participants in the fee-for-service system (P.L. 111-148). The feedback reports are a mechanism to provide transparent and comprehensible performance results, and to encourage clinicians to provide more efficient and higher-quality care (CMS, 2011e).

CMS also will be making the quality data for physicians and other health professionals available to the public through the Physician Compare website,⁷ analogous to Hospital Compare. The site currently allows anyone to search for a physician or other health care professional by specialty, type of professional, and location. To meet requirements of the Medicare Improvements for Patients and Providers Act, a list of professionals who satisfactorily reported PQRI measures for 2009 is also available on the site. CMS is required to implement a plan for making the performance data available to the public by January 2013.

National Quality Strategy

As documented in the most recent National Healthcare Quality Report (AHRQ, 2010), improvements in the quality of care have been disappointingly slow and have not yielded significant across-the-board changes in provider performance. For example, the widely studied Premier hospital demonstration, initiated in 2003, has not resulted in changes in mortality and

______________

⁶ Eligible providers include doctors of medicine, osteopathy, podiatry, optometry, oral surgery, dental medicine and chiropractic medicine, physician assistants, nurse practitioners, clinical nurse specialists, certified registered nurse anesthetists, anesthesiologist assistants, certified nurse midwives, clinical social workers, clinical psychologists, registered dieticians, nutrition professionals, audiologists, physical therapists, occupational therapists, and qualified speech- language therapists.

⁷ See http://www.medicare.gov/find-a-doctor/provider-search.aspx.

has called into question the value of using financial incentives in hospital pay-for-performance programs (see Box 4-1).

In March 2011, HHS (2011a) released the National Strategy for Quality Improvement in Health Care (National Quality Strategy), a congressionally mandated report with the first set of national aims and priorities to guide efforts to improve national, state, and local efforts and help reduce their administrative and quality reporting burdens. Building on the national strategy, CMS (2012a) released National Impact Assessment of Medicare Quality Measures, a congressionally mandated report on quality and efficiency measures that marks a shift for Medicare toward playing a more active role as a value-based purchaser. A technical expert panel will be convened to identify critical measures and areas for future such reports and will include a mix of measure types, including structural, process, outcome, patient experience of care, efficiency, care transitions, and system integration measures (CMS, 2012a). Taken together, these coordinated national efforts may begin to yield better results in terms of patient outcomes and also reduce the burden of reporting, which currently involves data collection from claims, assessment instruments, medical charts, and registries.

Another influence on quality reporting will be the increasing use of electronic health

records to collect and report quality measures. Authorized by provisions of the HITECH Act,⁸ CMS provides Medicare and Medicaid incentive payments for health providers who demonstrate “meaningful use” of certified electronic health record systems for such purposes as e-prescribing, exchange of clinical information at the point of care, and quality reporting (CMS, 2012b). Electronic reporting of performance data as a routine part of clinical care is expected to reduce administrative burdens, provide more accurate and timely data, and help to improve the quality of care.

As described in Chapter 3, some geographic variations in quality are known to occur, but extensive variation also occurs across settings even within the same geographies. There is no evidence to support using geographic adjustment to try to affect the quality of care, and the committee does not view geographic adjustment as an appropriate means to do so. Therefore, while the committee views quality reporting and accountability as essential to health care transformation, no recommendations in this report will address health care quality.

CURRENT PROGRAMS TO IMPROVE WORKFORCE SUPPLY AND ACCESS

This section examines evidence on the extent to which various programs and policies have influenced the ability of U.S. communities to attain adequate access to services appropriate to meet their health care needs. It begins with a brief conceptual section that identifies potential barriers to access and types of policy solutions designed to trigger workforce supply responses. It then reviews the evidence on the extent to which such programs and policies have been successful in improving access, and it discusses the ways in which Medicare payment policies may have also affected practitioner location and access to care.

Policies to Increase the Workforce Supply

Locations and practitioners differ in many important ways. Areas differ, for example, in the costs of inputs required to establish a practice and provide health care services, such as remodeling an office, hiring staff, and building a patient clientele. As the committee’s Phase I report indicated, it is very important to measure input cost variations accurately and incorporate them appropriately in setting fee-for-service payments to Medicare’s providers.

Practitioners also show variation in the importance they place on both financial and nonfinancial characteristics of their practice locations. On the financial side, income per professional in areas that are generally attractive may be lower than in those areas that are relatively unattractive, because salaries and/or bonuses may be higher to attract health professionals.

As following sections will discuss, financial incentives and considerations of rate of return on the medical school investment are an important part of the decision-making process for specialty and location choices for physicians (Nicholson, 2008), but they are not the only factors involved in these decisions (Phillips et al., 2009). For NPs and PAs, as discussed in Chapter 3, rural practice sites are more likely choices, but those are not reported in the literature on the impact of federal loan repayment programs.

______________

⁸ Health Information Technology for Economic and Clinical Health, part of the federal economic stimulus program authorized under the American Recovery and Reinvestment Act of 2009.

Physicians’ Location Decisions

In a standard economic framework employing several underlying assumptions about market pressures—such as no meaningful differences among geographic areas in the desirability of locations to those who are making a location decision or in per capita demand for care—increases in the supply of physicians in the United States as a whole should lead physicians to locate in smaller geographic markets, meaning rural areas (Newhouse et al., 1982). In other words, the geographic diffusion of physicians would be expected to respond to market forces (Ricketts and Randolph, 2007).

In fact, Rosenthal et al. (2005) demonstrated that diffusion to smaller areas occurred between 1979 and 1999 following a substantial increase in U.S. physician supply. They grouped physician specialties into four groups depending on the number of physicians in the field. In both years, the vast majority of communities with population sizes of 2,500 to 5,000 had at least one physician in group 1, general or family practice (86 percent in 1979 and 91 percent in 1999). However, in their group 2, less than half of the communities of this population size had at least one physician in general or family practice (23 percent in 1979 and 41 percent in 1999 for internal medicine, for example). The percentage of such communities having group 3 and 4 specialties was minuscule (e.g., 4 percent and 3 percent in ophthalmology, a group 3 field, and 1 percent in each year for neurology, a group 4 field). More generally, group 3 and 4 specialty physicians were located in larger versus less populated communities.

Using American Medical Association Masterfile data to study physician diffusion between 1981 and 2001, Ricketts and Randolph (2007) also found a small net flow from urban to rural areas, which they attributed to workforce programs that are intended to counter the normal market pressures for health professionals to locate in urban areas. One-third of the physicians they studied remained in the same urban or rural practice for most of their careers, and approximately one-quarter moved across county boundaries in any given 5-year period.

There are few studies on the role of financial incentives on physician location decisions. In a study of location decisions of psychiatrists, Frank (1985) obtained estimates of the elasticity of the short-run response to a fee change of 0.13 and 0.23, which are similar to the elasticity estimates reported for specialty choice in Box 4-2, and long-run response estimates of 0.96. However, it would take a long time for the long-run response to be realized.

In the short term, most physicians face the barrier of having to reinvest in practice building in a new location among other barriers to relocating, and hence are likely to stay where they are. The fraction of physicians who are willing to consider making a locational choice are largely limited to the pool of recent medical school graduates. At any point in time, this pool is small relative to total physician supply. If fee differences persist over time, there will be more graduating classes of physicians making location decisions and some less recent graduates may be willing to relocate as well.

An increasing number of medical school graduates are choosing to become salaried employees rather than establish their own independent practices, and it is not clear how much this trend will impact physician diffusion. Two studies have examined the effect of health maintenance organization (HMO) market penetration on physician location (Escarce et al., 1998; Polsky et al., 2000). Although not a direct measure of physician earnings, HMOs presumably have a negative effect on such earnings, both by negotiating lower fees and by reducing demand for care. The bottom line from these studies is that HMO penetration has a larger effect on location of younger physicians than it does on physician location decisions overall. This evidence is consistent with the view that the primary location changers are younger physicians.

Practitioner Preferences

Three types of strategies deal with variation in location in attractiveness: personnel selection, educational policies that seek to demonstrate the positive features of areas with certain characteristics, and policies to increase the attractiveness of location. Selection policies recognize inherent heterogeneity in preferences among individuals (Wilson et al., 2009). Such policies seek to recruit persons into training who for reason of their backgrounds or stated career orientations are likely to be willing to practice in areas that the majority find unattractive for personal or professional reasons. Educational policies provide exposures that are likely to make placements in such areas more likely to succeed. In a sense, they seek to change preferences rather than to take them as given as in the selection policies.

The third approach improves the professional attractiveness of areas. There are essentially two dimensions to nonfinancial attractiveness: professional and personal. From a professional perspective, connectedness with developments in medicine (e.g., the ability to enhance connectedness to professional expertise using telemedicine for videoconferencing with colleagues in other locations, good colleagues and facilities, and clinical backup) are attractive features. In addition, personal factors include quality of schools for children, availability of employment opportunities for spouses, recreational opportunities, and connectedness to the community, such as prior experience with that type of community.

Thus, health workforce policy needs to account for variations both in the characteristics of geographic areas and variations in the preferences of practitioners. A limited amount of empirical evidence gives some indication of the effectiveness of alternative policies, and that will be presented in a subsequent section.

Barriers to Entry

As previously discussed, geographic areas may differ in the cost of establishing a practice, such as the costs of becoming licensed and credentialed in a location, remodeling an office building, and building a practice. Variation in entry costs may reflect differences in policies among areas, such as in certificate of need (e.g., for an ambulatory surgery facility), openness of hospital medical staffs, and enforcement of no-compete provisions of employment contracts. If entry cost in an area is sufficiently high, practitioners may be unwilling to locate there, and policy interventions such as bonus payment programs may be needed to ensure adequate access.

Analytically, there is a difference in the potential effect of a public policy such as a loan forgiveness program, which offsets entry cost, and a reoccurring bonus which takes the form of a fixed-dollar annual subsidy, a bonus per unit of service, or a combination of the two. A one-time subsidy, such as a loan forgiveness program, encourages entry but will not provide a financial disincentive for exit if ongoing conditions are not favorable for the provider. In contrast, a fixed-cost subsidy that reoccurs annually provides a slight entry incentive and a more powerful disincentive for exit.

Thus, it is important to understand areas’ conditions of entry and disincentives to exit in order to devise effective strategies to enhance and maintain access. These are discussed further in the next section. In addition, Box 4-2 discusses the relationship between expected earnings, specialty choice, and practice location decisions.

EMPIRICAL EVIDENCE ON EFFECTS OF PUBLIC POLICIES TO IMPROVE GEOGRAPHIC DISTRIBUTION OF HEALTH CARE PRACTITIONERS

Limited evidence exists on the factors that determine location decisions, and it is almost entirely about physicians. In a comprehensive study of physician career choices by a research organization affiliated with the American Academy of Family Physicians, data surveying graduating medical students over a period of nearly 20 years were brought together to analyze student characteristics and training influences that might influence their choices of specialty and geographic practice location (Phillips et al., 2009). The specific outcomes studied were practicing in primary care (family medicine, general internal medicine, or general pediatrics), a rural community, a health center (either a Federally Qualified Health Center or rural health center), an underserved area, or ever having served in the National Health Service Corps (NHSC). Data sources included historical Title VII⁹ training files, cross-sectional data about current specialties and practice locations, and a 5-year cross-section of service in Rural Referral Centers and Federally Qualified Health Centers.

In general, the study found that public and rural training programs produced a higher proportion of primary care, rural, and health center physicians than medical school programs that did not have a programmatic emphasis on underserved populations. Students in the study

______________

⁹ Title VII of the Public Health Service Act provides for the National Health Service Corps and other programs to expand the geographic, racial, and ethnic distribution of the health care workforce.

who chose to accept debt obligations such as loan repayment were more likely to practice in primary care and rural and other underserved areas. Other predictors of choosing careers in primary care and rural and underserved areas were being born in a rural area, being interested in serving underserved or minority populations, or having had Title VII experience in medical school and training experiences in rural or inner-city areas. Title VII exposure during residency increased the likelihood of serving in the NHSC and in shortage areas, but not in primary care or rural practice. In addition, men were less likely to choose primary care, and women were less likely to choose rural practice.

Students with no debt and no scholarships (either NHSC or armed forces) were the least likely to later practice in primary care, a rural area, or a CHC. The authors concluded that addressing the income gap and its consequences will require changes in the way training is financed and the settings in which training is provided for physicians. There is no comparable study for other health professionals, but these choices may be unique to the physician labor market because of the generally higher incomes in the specialty and subspecialty categories. That may change as NPs and PAs increasingly begin to specialize because of the financial incentives associated with specialty care.

Further review of the empirical evidence in this report examines programs designed to reduce the cost of entry into underserved areas, direct public provision of services, and programs designed to encourage continued practice in underserved areas, including increased professional connectedness and targeted training. The committee looked for literature on the impact of Medicare’s payment policies on health professionals’ location decisions. Although payment policies may affect location and policy makers should consider the possibility that their decisions could affect health professionals’ location decisions, the committee found no conclusive empirical evidence specifically linking payment policies to such choices.

Policies Affecting the Net Cost of Entry—Outcome
Evaluations of National Health Service Corps

The NHSC was created by the Emergency Health Personnel Act of 1970 (P.L. 91-623).¹⁰ This legislation authorized the U.S. Public Health Service to assign commissioned officers and federal civil service personnel to practice in shortage areas. In 1972, Congress passed the Emergency Health Personnel Amendments authorizing scholarships to support health professions education in return for a minimum of 2 years of service in shortage areas designated by the agency. During the 1980s, the NHSC implemented the Loan Repayment Program, which substantially increased the number of NHSC field personnel.

Much of the literature evaluating the performance of the NHSC has focused on retention of health professionals in NHSC-designated shortage areas, and not on factors related to their recruitment. Based on survey data collected from NHSC and non-NHSC physicians practicing in similar settings during mid-1979 through year-end 1981, Pathman et al. (1992) found substantially lower retention rates for NHSC than for non-NHSC physicians. When these physicians were resurveyed in 1990, 12 percent of NHSC physicians remained in the practice they were in during the previous survey or within 24 km of this practice versus 39 percent for non-NHSC physicians. Nearly one-third (29 percent) of the former and slightly more than half (52 percent)

______________

¹⁰ The history of the NHSC is described in several sources. See, e.g., Politzer et al., 2000.

of the latter were in nonmetropolitan practices. The decline in retention was higher at the time the service obligation was completed, but the rate of drop off slowed thereafter.

Subsequently, Pathman et al. (1994a) analyzed data on cohorts of NHSC and a comparison group of non-NHSC physicians over the period 1987–1990, finding that retention rates¹¹ were about the same for NHSC as for the comparison group immediately after the NHSC obligation was satisfied, but beginning at 3 years postobligation, the NHSC physicians were less likely to remain at their practice sites. Some NHSC physicians reported negative experiences with their placements, which was one reason for low retention. Also, NHSC physicians were less likely to have been raised in rural areas than were physicians in the non-NHSC comparison group. Pathman et al. (2006a) focused on an underrepresented minority cohort of NHSC physicians, finding that the 1-year retention rate of these physicians was not statistically different than for other NHSC physicians.

Rosenblatt et al. (1996) argued that the concept of retention in the earlier studies was too narrow. They defined five retention measures: (1) physician remained at same rural site; (2) physician remained in same county; (3) physician practiced in remote rural county; (4) physician practiced in other rural county; and (5) physician now practiced in urban site (CK). They reported that three-fifths of physicians had left their original site (categories 2–5). Using a broader definition of retention similar to Rosenblatt and colleagues, Porterfield et al. (2003) found three factors were significantly associated with a higher probability of retention: older; higher initial desire to serve the underserved on an index ranging from 1 to 5; and final salary in the NHSC. The result for salary was that each $10,000 increase in final salary raised the probability of retention by 11 percent.

Data collected during the past 2 decades documented increases in NHSC retention rates (Politzer et al., 2000). In FY 2009, the retention rate was 76 percent, and this rose to 82 percent in 2011 (HRSA, 2012b). This may be due in part to changes implemented by the NHSC program, but it also could reflect changes in how retention is measured. HRSA is working on a retention strategy, and close to 60 percent of NHSC sites report that they have recruitment and retention plans (HRSA, 2012b). It is difficult to know whether the results of these evaluations are determined by the underlying physician selection process or whether there is something about the NHSC experience that promotes or deters retention. The result on salary, for example, might represent an effect of increased salary on retention decisions or simply the fact that more productive NHSC physicians are more likely to continue to work in shortage areas, or some combination of the two.

Another issue relates to whether or not the presence of NHSC physicians in an area reduces the number of non-NHSC physicians. A decrease might be expected since income per physician may fall with entry of new physicians, whether such physicians are NHSC or not affiliated with the NHSC. In investigating this issue, Pathman et al. (2006a) held other factors constant and found that more NHSC physicians led to higher growth in non-NHSC physicians in the area. A third unmeasured factor may account for both NHSC physician supply and the growth of non-NHSC supply.

Pathman et al. (2006b) compared growth in primary care physicians per 10,000 population during 1981–2001 in areas with an NHSC presence compared to those without NHSC clinicians. The physician-to-population ratio grew at more than twice the rate in the NHSC areas as opposed to the non-NHSC areas, which suggests that NHSC presence may make a positive

______________

¹¹ Defined as still working in rural practice at a 9-year follow-up.

contribution to the workforce in those areas. The impact of NHSC on NP and PA decisions to stay in underserved areas is unknown, because no data were found on these groups of practitioners.

Process Evaluations of the NHSC

Pathman et al. (1994a) surveyed physicians in the NHSC and a comparison group of physicians to learn about the quality of their experiences. They found that NHSC physicians often were motivated to practice in areas characterized as “underserved,” a result consistent with evidence from other studies (see, e.g., Phillips et al., 2009; Probst et al., 2003). Yet, Pathman and coauthors suggested that the NHSC could have done more to accommodate the preferences of NHSC personnel and their families, and this has been an issue in NHSC retention. The Office of Inspector General of HHS (OIG, 1994) conducted an evaluation of NHSC processes. A deficiency documented by the Office of Inspector General was the inadequacy of the matching process. Furthermore, prospective NHSC physicians lacked information on what was involved in practicing in underserved areas. Two GAO studies (1995, 2001) concluded that there was an imbalance of placements with some sites receiving excess numbers of placements and others receiving none. In recent years, NHSC has been making priority placements of personnel to ensure that assignments are made to areas with the greatest need (Salinsky, 2010), but funding shortfalls mean that large number of unfilled vacancies remain.

Other Programs That Reduce Net Entry Cost: State
Scholarship, Loan Forgiveness, and Related Programs

Pathman et al. (2000) surveyed state programs in 1996 that satisfied the following criteria: (1) provided financial assistance to medical students, physicians, physician assistants, and advanced practice nurses; (2) conditional on providing service in a designated medically under-served area ("service for support programs"); (3) relied on public and/or private philanthropic financial support paid to an individual, educational, or financial entity; and (4) were statewide rather than in a particular locality within a state. They identified programs in 41 states consisting of scholarships (n=29), loans (n=11), resident support (n=5), loan repayment (n=29), and direct financial incentive programs (n=8). Scholarship programs have an expectation that students will provide service upon graduation with substantial financial penalties for those who do not. Loan programs have service requirements but offer the option of repaying the loans in lieu of service at market interest rates. Direct financial incentive programs offer incentives at about the time the health professional is to enter practice but provide unrestricted funds. Resident support programs provide financial assistance to residents with 1- to 2-year service requirements at the end of the residency.

The aggregate number of health professionals receiving subsidies was small; an estimated 1,215 practitioners signed initial contracts and 1,676 (of whom four-fifths were physicians) were working to fulfill their service obligations. For physicians, financial support ranged from $3,000 to $38,000 per annum with service obligations from 1 to 60 months. The authors did not perform an evaluation of program outcomes.

More recently, Pathman et al. (2004) conducted an evaluation of selected state program outcomes based on surveys sent to state-obligated physicians and a comparison group of physicians in 1998–1999. The loan programs had the lowest mean service completion rate (44.7 percent, followed by the scholarship programs at 66.5 percent). The remaining types, which

in contrast to the loan and scholarship programs incentivized physicians much nearer to the time they first entered practice, had service completion rates of over 90 percent. Physicians in the obligated group were more likely to be practicing in areas with greater underservice than were physicians in the comparison group. Physicians in the former group were more frequently satisfied with their work and practices than were those in the latter group. Over 90 percent of those in the obligated group said that they would be likely to enroll in the program again. That is, they had no regrets. Retention was higher among the obligated than the nonobligated group, although the difference diminished as the length of the follow-up period increased. By year 8 following placement, retention rates were 55 percent for the obligated versus 52 percent for the nonobligated group.¹² In evaluation of similar programs in a single state (West Virginia), retention rates were slightly over 50 percent at 10 years for the obligated and between 60 and 70 percent for the nonobligated comparison group.

J-1 Visa Waiver Program

The proportion of physicians practicing in the United States who graduated from medical schools outside of the United States and Canada, termed international medical graduates (IMGs), has grown appreciably in the last 5 decades and now represents approximately one out of four practicing U.S. physicians (AMA, 2006). This growth is attributable to a variety of factors, including changes in U.S. immigration policies as they apply in general and to health professionals in particular and expanded capacity in U.S. residency programs that exceeds the level that can be filled by U.S. medical school graduates. The J-1 Visa Waiver Program allows non—U.S. citizen IMGs to enter the United States for educational purposes under the “alien physician program.” Upon completing their postgraduate residency programs, physicians holding a J-1 visa must return to their countries of origin for 2 years before becoming eligible to return to the United States. However, if they agree to work in a designated position in an HPSA, they can apply for a waiver of this requirement provided that they work in a HPSA for 3 years. Thereafter, they are free to practice in nonshortage areas.

Current sources of J-1 visas are the HHS and the Conrad-30 (or State 30 waiver) Programs (CK). Under the latter, each state is allowed 30 visa waiver slots annually to administer through its state health department. An alternative to a J-1 is an H-1B visa. The J-1 is for training while the H-1B is for “temporary specialized workers.” The latter is more flexible in that there is no requirement of return to the home country after training. On the other hand, it imposes a requirement that the spouse not work and takes more time to process than the J-1 does.

In late 1999, over 2,000 IMGs with J-1 visa waivers were practicing in shortage areas compared to 1,356 physicians in the NHSC (GAO, 2001). According to data from the Educational Commission for Foreign Medical Graduates, the number of IMGs with J-1 visas in 2006-2007 declined by more than 5,000 individuals compared with the numbers from 1996-1997 (Boulet et al., 2006; Croasdale, 2008).

Conceptually, the J-1 Visa Waiver Program reduces the entry cost on IMGs for practicing in the United States. Assessments of the effect of IMGs in general and the J-1 Visa Waiver Program in particular have been very limited. A public policy issue of general concern is whether entry of IMGs has contributed on balance to a reduction in rates of underservice in the United States.

______________

¹² In multivariate analysis, which controlled for demographic characteristics and physician specialty, difference in retention rates was below conventional statistical significance levels (p=0.08).

Mick et al. (2000) investigated this issue empirically and found that IMGs are more likely than U.S. medical graduates to practice in underserved areas. This finding was supported by Mertz et al. (2007) but not by Fink et al. (2003), reflecting differences in study methods.

In Wisconsin, a survey of physicians with J-1 visas indicated general satisfaction with the medical community and with the care provided (Crouse and Munson, 2006). However, physician lack of integration into the community was perceived as a problem. J-1 Visa Waiver Program participants who did not integrate into the community were significantly less likely than those without J-1 waivers to stay in the community 3-4 years after initial location. Among the reasons cited for lack of integration were unmet cultural and religious needs (Crouse and Munson, 2006).

Multivariate analysis in a study of J-1 Visa Waiver Program participants conducted in the state of Washington (Kahn et al., 2010) revealed no statistically significant predictors of retention. However, from open-ended questions about how the program could be improved, nearly two-fifths of physician respondents said that they felt employers could have shown more respect (such as treating them the same as they did nonwaiver employees), could have offered more support (such as with legal help with visa applications), and could have offered more competitive compensation (at market rates).

Targeted Training

There are large differences among medical school graduates in the propensity to practice in rural areas (Rosenblatt, 2010). While these differences plausibly reflect self-selection as well as effects of exposure to rural practice during the course of medical education, evidence seems clearest for selection. In a qualitative study based on 22 in-depth interviews with rural- and urban-raised physicians in northeastern California and northwestern Nevada, Hancock et al. (2009) identified types of exposures, most outside of the medical education process, that led physicians to choose to practice in rural areas. Another important influence on rural practice recruitment is having been brought up in a rural area (Blue et al., 2004; Hancock et al., 2009; Phillips et al., 2009).

As for the role of medical education, several points in a health professional’s decision-making process are relevant to the ultimate goal of improving availability of health professionals in underserved areas. These include choice of specialty (e.g., Quinn et al., 2011, Rabinowitz et al., 1993), initial location decision after training—the recruitment dimension, and decision to remain in an underserved area after practice obligations are satisfied—the retention dimension. Subspecialists are less likely to locate in a rural area or even an inner city because the market size is insufficient to support the practice (e.g., Ricketts and Randolph, 2007; Rosenthal et al., 2005).

Factors influencing recruitment are distinct from those affecting retention in a geographic area underserved by health professional personnel. Recruitment reflects such factors as exposures during childhood and adolescence, among other factors. By contrast, retention is from a self-selected minority of personnel who have been willing to practice in such areas at all. For example, self-reported preparedness for rural practice and small-town living was positively associated with recruitment but not retention in one study (Pathman et al., 1992).

Pathman et al. (1994b) inquired whether retention in rural practice is of longer duration for public medical school than for private medical school graduates, for those who participated in community hospital-based residencies, and for physicians who had participated in rural rotations as medical students and residents. In 1980 to 1982, the authors identified 464 rural practices

in the United States that received some kind of external subsidy, such as from the NHSC, CHC funds, and state and local governments. In 1990, they conducted a follow-up study of 412 physicians identified at baseline in primary care fields. Respondents were asked about rural training experiences as well as more general questions about their medical education. They found that among non-NHSC physicians practicing in subsidized rural practices at follow-up, retention duration could not be predicted by ownership of medical school, training in community hospital-based residencies, or participation in rural training experiences as medical students or residents. Among NHSC physicians in such practices at follow-up, graduates of private medical schools remained in these practices longer than their counterparts from public medical schools did. The authors explained that a physician’s rural upbringing is a predictor of recruitment to rural areas (Pathman et al., 1992), but it is not a predictor of retention.

A study for the Josiah Macy Foundation (Phillips et al., 2009) analyzed career decisions using almost 20 years of survey data from graduating medical students related to student factors (e.g., demographics, values), curriculum factors (medical school training), debt, and institutional factors (public or private). While rural birth and training, attending a public medical school, and attitudes about assuming debt were important predictors for rural primary care careers, interest in underserved populations nearly tripled the likelihood of practicing in an FQHC and made service in the NHSC eight times more likely.

These findings suggest that recruitment and retention efforts should give more weight to personal interests in working with the medically underserved, beginning with acceptance to medical school and continuing through mentoring programs, internships, and residencies. Clearly, the relationships between personal background, personal interests, training experiences, exposures, and career decisions are more complex than is commonly believed and need to be better aligned.

Medicare Implications of Findings

Various federal and state workforce policies have been implemented over the past decades with the intent of increasing production of health professionals determined to be in short supply, to support clinical training—especially in locations serving underserved populations—and to encourage professionals to practice in underserved areas. The availability of evidence to determine the effectiveness of these programs varies greatly by program, from none to numerous studies.

While residents of underserved communities probably have benefited from the public programs on balance, it might be said that the programs have not been implemented on a sufficient scale to have had meaningful effects or to have had meaningful effects in some areas but not all. For example, the funding levels for the NHSC have never been adequate to support the number of clinicians who would be required to fill all of the vacancies (Salinsky, 2010). There are many such programs with a relatively small investment per program, especially relative to the size of the U.S. health care sector overall and public programs such as Medicare and Medicaid in particular.

There has been little theoretical and empirical analysis of geographic markets for health professionals, and there is a paucity of relevant studies in the peer-reviewed literature. In part, this reflects the lack of availability of general grant support for research on health care supply, organization, and financing. Moreover, there is no consensus about the endpoints that should be used in program evaluation. For example, much of the research on effectiveness of loan

forgiveness programs has been on physician retention following fulfillment of the service obligation, using different numbers of years of continuing service to indicate success. However, the effect of physician turnover on patient care is unknown. In general, while a variety of programs attempt to enhance recruitment, retention, or both, we know very little about which types of policies are most successful in improving access in underserved areas.

CMS and HRSA are both HHS agencies, but policies to enhance access to primary care services have not been jointly developed.¹³ Indeed, Medicare payment policies that promote specialization and a large income gap between primary care practitioners and specialists have likely worked at cross-purposes with the objectives of Public Health Service programs to improve access in underserved areas. As Medicare is the single largest insurance program in the world, the incentives produced by its payment policies may well dominate many decisions made by health care providers throughout the United States.

DELIVERY SYSTEM REFORM INITIATIVES TO PROMOTE
ACCESS AND QUALITY OF CARE

The growing costs of health care, the demographics of the aging population, and the challenges of managing the care of older adults with multiple chronic health conditions are driving system reform and innovations in health services delivery. Among these are changes intended to improve access to primary care services and to improve the coordination of care as mechanisms for improving access and health care outcomes.

There are many emerging conceptual models of coordinated care, including accountable care organizations, transitional care, medical homes, and others (CMS, 2011g; CRS, 2010; Friedberg et al., 2010; Naylor and Kurtzman, 2010; Pohl et al., 2010). What these models have in common are (1) the essential role of primary care services in ensuring that the care is coordinated and provided in the appropriate setting and level of care, including the type of clinician who provides services, and (2) the need for payment reform to support the policy goal of improved care coordination.

Over time, many types and models of care may be shown to improve clinical outcomes and population health. Because Medicare is the largest payer, many of the new models are intended to move Medicare payments away from fee-for-service payment by providing financial incentives for shared risk through bundled payment options. It is hoped that these models will not only improve efficiency but also provide better integration and more coordinated care for beneficiaries (Guterman et al., 2009). The committee expects that these models may have workforce implications, by increasing the demand for primary care services provided by NPs and PAs.

Accountable Care Organizations (ACOs)

ACOs are defined by CMS as “groups of physicians, hospitals, and other health care providers who come together voluntarily to give high-quality coordinated care to the Medicare patients they serve” (CMS, 2012c). The goal of ACOs is to improve quality of care for Medicare beneficiaries by coordinating care among practice settings (e.g., hospitals, physician groups, and skilled nursing facilities), which helps ensure that patients get the appropriate level of care and that unnecessary duplication of services, medical errors, and hospital readmissions are

______________

¹³ Jurisdiction over Public Health Service programs and Medicare is also exercised by different committees in the U.S. Congress.

reduced (CMS, 2012c). CMS has established two ACO payment programs to provide financial incentives for Medicare-enrolled providers who come together to form an ACO: the Medicare Shared Savings Program and the Advance Payment Model.

Medicare Shared Savings Program

Shared savings is an approach to improving the value of health care by promoting accountability, requiring coordinated care, and encouraging infrastructure investments such as electronic health records and broadband to enable the secure exchange of clinical information across settings in real time (CMS, 2012d). Other investments may include hiring new nurse care managers and other personnel to provide better continuity of care across clinical settings. Participants agree to lower the cost of health care while meeting identified performance standards by sharing resources and care in a coordinated manner.

CMS is implementing two shared savings models: one-sided and two-sided shared savings models. Under the one-sided model, providers only share the savings; in the two-sided model, providers share the savings and the losses if there is a growth in costs (CMS, 2012d). CMS will develop a target level of spending for each participating ACO. Depending on the quality of their performance, those under the one-sided model will receive up to 50 percent in savings and those under the two-sided model will receive up to 60 percent in savings (CMS, 2012d).

According to the ACA, NPs are authorized to be ACO professionals but are excluded from assignment of patients for this program; that is, patients must be assigned to and cared for by primary care physicians. Therefore, patients cannot choose a NP for their primary care provider under the current ACA. While this does not prevent nurse practitioners from joining an ACO, it does prevent their patients from being assigned to them directly. At this early stage, it is not clear how third-party payers will respond to this.

Pioneer ACO Model

The pioneer ACO model is designed to support organizations that already have experience operating as ACOs or in similar arrangements providing coordinated care to Medicare beneficiaries at a lower cost to Medicare. It is designed to allow them to move more rapidly from a shared savings payment model to a population-based payment model and to work in coordination with private payers to provide better care for beneficiaries (CMS, 2012e).

The first 32 pioneer ACO organizations were announced in December 2011 after a lengthy and competitive process (CMS, 2012f) and collectively provide care for about 860,000 beneficiaries. They include primarily physician-led organizations and include health systems in urban and rural areas in 18 states and various geographic regions of the country. The pioneer initiative is operated by the CMS Innovation Center and tests a shared savings and shared losses payment arrangement with higher levels of reward and risk than the rest of the Shared Savings Program.

Advance Payment Model

The advance payment ACO model is open only to two types of organizations participating in the Shared Savings Program: (1) ACOs that do not include any inpatient facilities and have less than $50 million in total annual revenue; and (2) ACOs in which the only inpatient facilities are critical access hospitals and/or Medicare low-volume rural hospitals and have less than $80 million in total annual revenue (CMS, 2012g).

The advance payment model is designed to provide financial support to organizations by improving their access to capital, including rural and physician-owned organizations. The first five awardees were announced in April 2012, and additional organizations were announced in July 2012 (CMS, 2012g).

Future Success of ACOs

The first 27 ACOs were selected in April 2012, including the five that are participating in the advance payment model, and they will be coordinating care for nearly 375,000 beneficiaries (CMS, 2012h). Of the 27 organizations, 21 are physician-led, even in those ACOs that involve hospitals (Fiegl, 2012). While the goal of all ACOs is to improve quality at lower costs, it is not yet clear how many new employment opportunities for NPs and PAs will emerge. Regional differences in staffing may be observed based on the current number of NPs and PAs already practicing in local markets.

ACOs will have many operational challenges, including the need to change beneficiaries’ expectations about how they receive their care, beneficiaries’ potential unwillingness to share their personal health information with other providers within the ACO network, and difficulties establishing secure health information exchanges to support the exchange of that information (Gold, 2012).

As hospitals and physician groups are consolidating to form ACO networks, there is concern that mergers and provider consolidation could increase market share enough to provide more leverage in negotiations with private insurers, thus driving up health care costs—having the opposite effect from the one intended. There is also concern that providers may take on more financial risk than they can handle, that the quality standards are too rigorous, that the expense of quality reporting through chart reviews or surveys for those who do not have electronic reporting will be overly burdensome, and that the potential savings will be too low in relation to the upfront investments needed (Ginsburg, 2011).

From a workforce perspective, one of the more controversial aspects of ACOs is that beneficiaries are attributed to ACOs on the basis of which primary care physician provided a plurality of their primary care services. It will not be clear for some time how nurses, NPs, and PAs will contribute to care coordination across organizations and settings, one of the main goals of the ACO program.

Value-Based Purchasing

CMS will begin the hospital value-based purchasing program in FY 2013, requiring that a certain percentage of Medicare hospital payments be based on hospital performance.¹⁴ The goal of the value-based purchasing program is not only to incentivize hospitals to improve the quality of care they provide, but also to reward hospitals based on the extent of their quality improvement.

The hospital performance standards for the value-based purchasing program are based on 12 clinical process measures and 8 patient experience measures and were announced in 2011

______________

¹⁴ CMS will use a linear exchange function to compute the percentage of value-based incentive payment earned by each hospital. CMS will notify each hospital of the exact amount of its value-based incentive payment by November 1, 2012 (CMS, 2011b).

(CMS, 2011i). Each hospital will be scored based on achievement and improvement ranges for each of the 20 quality measures (CMS, 2011i). Achievement scores are to be based on how much a hospital’s current score differs from all other hospitals’ baseline period performance. Hospitals will be awarded achievement points if they fall within the range of the 50th percentile of hospital scores or higher (CMS, 2011a). Improvement scores will be based on how much a hospital’s current performance changes from its own baseline period performance, and hospitals that meet or exceed the performance standards will receive a value-based incentive payment. Future measures for consideration are posted on the CMS Hospital Compare website (CMS, 2012i).

Incentive Payments for Coordination of Care

New bonus payments for primary care practitioners are funded under the Affordable Care Act to improve coordination of care. Beginning in the summer of 2012, a Medicare pilot project, referred to as the Comprehensive Primary Care Initiative, will pay primary care practitioners in five to seven markets a bonus of $20 per month per patient for helping patients to coordinate care with specialists, develop personalized care plans, and deliver preventive care and health education (HHS, 2011b). This initiative is based on evidence from previous pilot projects and other studies that show primary care services help to prevent and reduce the use of more complex and more expensive settings, often with better patient experiences and outcomes (e.g., Friedberg et al., 2010).

Among the most vulnerable times for patients—especially those with chronic conditions—are the times when they are transitioning from one clinical setting to another, such as returning home after a hospitalization or nursing home stay. The movement of patients from hospitals and nursing homes to their homes and back to clinical settings has been estimated to cost Medicare approximately $15 billion per year. Studies have shown that up to 34 percent of beneficiaries experience adverse events and/or are rehospitalized at those transition times, which are also known as “handoffs” (Naylor et al., 2011). New approaches to improve care integration across episodes and settings of care are referred to as transitional care. These practices are distinct from care coordination in that they focus on critical transition periods, are time limited, designed to avoid preventable hospitalizations, and supported by a robust body of evidence that confirms their benefits (Coleman et al., 2006; Naylor et al., 2011).

POLICY CONSIDERATIONS AFFECTING WORKFORCE DISTRIBUTION

In 2009, nearly 1 million health professionals were participating in fee-for-service Medicare (MedPAC, 2011), but there are no agreed-upon national targets for the supply of practitioners by type. A major source of concern among policy experts involves questions about whether the balance and coordination of primary care services and specialty care are meeting the needs of beneficiaries for chronic care management and other primary care services (Goodman and Grumbach, 2008; Naylor and Kurtzman, 2010).

Variation in Payment Policies for Medicare Providers

In addition to payment for physician services, Medicare also pays for NP or PA services. These payments are paid at the same rate paid to physicians only if the services are deemed to be “incident to” physician services. This means that the service cannot be billed to Medicare

by the NP unless a physician has seen the patient previously for the particular diagnosis being addressed at the visit (i.e., so the NP service is considered to be “incident to” the physician’s management). In practice, this means an NP may not see a new patient or an existing patient with a new diagnosis if billing “incident to” a physician. In settings where these practitioners are salaried employees paid by the supervising physicians, the payment benefits accrue to the physicians as employers, and in settings where NPs are paid by hospitals, any payment benefits accrue to hospitals.

To help expand access to primary care services, an incentive payment of 10 percent of the amount is paid quarterly under the Physician Fee Schedule for primary care services provided by primary care physicians (defined as those trained and practicing in family medicine, internal medicine, pediatrics, and geriatrics), NPs, clinical nurse specialists, and PAs (CMS, 2011b,f).

Along with differences related to the degree of practitioner autonomy and the legal right to practice independently, payment differentials between primary care and nonprimary care physicians and between physicians and other health professionals are controversial. While physician organizations often maintain that their longer and more extensive medical training makes them more qualified than NPs or PAs to diagnose and treat patients and that their payment should be commensurate with their experience, others focus on the need to expand primary care services through (1) the use of the “full primary care workforce” (Fairman et al., 2011; IOM, 2010; Pohl et al., 2010) and (2) consideration of the outcomes of care from all health care professionals (Newhouse et al., 2011).

National Workforce Policy and Data Gaps

As discussed previously in Chapter 3 and in this chapter, the committee’s review has been made more difficult by serious data gaps, including a lack of consistent methodologies for workforce studies, conflicting findings, and lack of research attention to many members of the health care workforce who provide care for Medicare beneficiaries. For example, as mentioned in Chapter 3, there is no single source of accurate, up-to-date information on the current numbers of practicing health professionals for all of the major professions, and many of the sources of data have biases or other flaws in their data collection methods or reporting. Another problem the committee encountered is the lack of current evaluation research on training programs, due to several years of underfunding for evaluations specifically and workforce programs generally.

As a result of the lack of clear, consistent data for many areas in its charge, the committee sought to draw conclusions from the limited evidence using a consensus process. Here too lay challenges, in that there were differences of opinion about how to distinguish between a shortage (supply) and a distribution problem, or how to describe the workforce implications for care coordination—a significant problem for beneficiaries in terms of access and quality of care, but not an area in which clear guidelines or practice patterns can be identified. The committee concluded that many of the complexities and contentiousness of workforce policies and programs are influenced by market-based factors such as competition for patients in local areas, as well as by professional levels and types of training and differing views about scope of practice.

Early in their deliberations, the committee members recognized the complexity of these issues and discussed the importance of creating and funding a new, independent body with representatives of different viewpoints that would help to prioritize workforce policy choices at a national level. They reviewed the authorizing language for a Health Care Workforce Com-

mission in the Affordable Care Act and agreed that that commission or an entity similar to it is called for. No single agency within HHS currently has the authority or the resources needed to provide a comprehensive, objective view of the nation’s workforce needs, and an independent body would be best suited to advise on ways to improve coordination across federal agencies.

FINDINGS

The committee members sought to identify the best available evidence to help them determine what policies and programs have been most effective in improving access to hospital and clinical services, improving quality of care, and helping to increase the supply of practitioners as well as influence their distribution across the country. The evidence review was hampered by the lack of evaluation, mixed results, and methodological challenges in finding studies that included geographic comparisons.

In developing their findings and recommendations, committee members developed a consensus on what the evidence base suggests. They also identified some promising new areas of policy and program development where changes in payment policies have the potential to expand beneficiaries’ access to care.

1.  Medicare policies intended to preserve beneficiary access to hospital care may not be efficiently targeted.

2.  The effectiveness of bonus payments to improve Health Professional Shortage Areas (HPSAs) has not been adequately evaluated, and it will take time before there is enough information and experience to evaluate other recently introduced bonus payments.

3.  Medicare’s payment policies related to quality of care are important, but there is no evidence that geographic adjustment is related to quality of care.

4.  Current information on public programs related to workforce is inadequate to assess whether current needs are being met. Consistent national data on workforce distribution and independent evaluations of public programs pertaining to distribution are lacking, and there are no nationally accepted distribution targets.

5.  Evidence suggests some success of federal loan repayments in placing practitioners in underserved areas. Retention rates appear to be comparable to retention of other practitioners in similar areas without special programs, but more studies with consistent definitions of retention are needed for comparison.

6.  New payment models are being introduced to encourage providers to improve care coordination through team-based approaches, but the extent to which an increased emphasis on care coordination will provide new opportunities for nurse practitioners and physician assistants to practice to the full extent of their educational preparation is unclear.

REFERENCES

AHRQ (Agency for Healthcare Research and Quality). 2010. National healthcare quality report. Rockville, MD: AHRQ.

AMA (American Medical Association). 2006. Physician characteristics and distribution in the U.S., 2006 ed. Chicago, IL: AMA.

Bazzoli, G. J. 1985. Does educational indebtedness affect physician specialty choice? Journal of Health Economics 4:1-19.

Berenson, R. A., R. A. Paulus, and N. S. Kalman. 2012. Medicare’s readmissions reduction program: A positive alternative. New England Journal of Medicine 366:1364-66.

Blue, A. V., A. W. Chessman, M. E. Geesey, D. R. Garr, D. H. Kern, and A. White. 2004. Medical students’ perceptions of rural practice following a rural clerkship. Family Medicine 36(5):36-40.

Boulet, J. R., J. J. Norcini, G. P. Whelan, J. A. Hallock, and S. S. Seeling. 2006. The international medical graduate pipeline: Recent trends in certification and residency training. Health Affairs 25(2):469-477.

Chan, L., L. G. Hart, T. C. Ricketts, and S. K. Beaver. 2004. An analysis of Medicare’s incentive payment program for physicians in Health Professional Shortage Areas. Journal of Rural Health 20(2):109-117.

Chin, M. H., S. B. Auerbach, S. Cook, J. F. Harrison, J. Koppert, L. Jin, F. Thiel, T. G. Karrison, A. G. Harrand, C. T. Schaefer, H. T. Takashima, N. Egbert, S. C. Chiu, and W. L. McNabb. 2000. Quality of diabetes care in community health centers. American Journal of Public Health 90:431-434.

Chin, M. H., M. L. Drum, M. Guillen, A. Rimington, J. R. Levie, A. C. Kirchhoff, M. T. Quinn, and C. T. Schaefer. 2007. Improving and sustaining diabetes care in community health centers with the Health Disparities Collaboratives. Medical Care 35(12):1 123-1 125.

CMS (Centers for Medicare & Medicaid Services). 2009. Fiscal year 2009 quality measure reporting for 2010 payment update. http://www.cms.gov/HospitalQualityInits/downloads/HospitalRHQDAPU200808.pdf (accessed August 10, 2011).

CMS. 2011a. Fact sheet: CMS issues final rule for first year of hospital value-based purchasing program. http://www.cms.gov/apps/media/press/factsheet.asp?Counter=3947 (accessed August 11, 2011).

CMS. 2011b. HPSA/PSA bonus payments overview, https://www.cms.gov/hpsapsaphysicianbonuses/ (accessed August 10, 2011).

CMS. 2011c. Hospital quality initiative: Outcome measures. http://www.cms.gov/HospitalQualityInits/20_OutcomeMeasures.asp#TopOfPage (accessed August 11, 2011).

CMS. 2011d. Hospital quality initiatives: Hospital inpatient quality reporting program. http://www.cms.gov/HospitalQualityInits/08_HospitalRHQDAPU.asp#TopOfPage(accessed August 11, 2011).

CMS. 2011e. Medicare FFS physician feedback program/value-based modifier. https://www.cms.gov/PhysicianFeedbackProgram/01_Overview.asp#TopOfPage (accessed August 9, 2012).

CMS. 2011f. Medicare learning network: Federally qualified health centers. http://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNMattersArticles/Downloads/SE1039.pdf (accessed August 15, 2012).

CMS. 2011g. Medicare physician group practice demonstration. http://www.cms.gov/DemoProjectsEvalRpts/downloads/PGP_Fact_Sheet.pdf (accessed August 17, 2011).

CMS. 2011h. Medicare program; hospital inpatient prospective payment systems for acute care hospitals and the long-term care hospital prospective payment system and FY 2012 rates; hospitals’ FTE resident caps for graduate medical education payment. Federal Register 42 http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/FY-2012-IPPS-Final-Rule-Home-Page.html/ (accessed August 15, 2012).

CMS. 2011i. Open door forum: Hospital value-based purchasing, fiscal year 2013 for beneficiaries, providers, and stakeholders. https://www.cms.gov/Hospital-Value-Based-Purchasing/Downloads/HospVBP_ODF_072711.pdf (accessed August 11, 2011).

CMS. 2011j. Physician quality reporting system: Statute regulations program instructions. https://www.cms.gov/PQRS/05_StatuteRegulationsProgramInstructions.asp#TopOfPage (accessed August 9, 2012).

CMS. 2012a. National impact assessment of Medicare quality measures. Prepared by Health Services Advisory Group, Inc. HHSM-500-2011-AZ10C. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/QualityMeasures/Downloads/NationalImpactAssessmentofQualityMeasuresFINAL.PDF (accessed May 22, 2012).

CMS. 2012b. EHR incentive programs. https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/index.html (accessed May 28, 2012).

CMS. 2012c. Accountable care organizations. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ACO/index.html?redirect=/ACO/ (accessed May 27, 2012).

CMS. 2012d. Shared savings program. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ACO/index.html?redirect=/ACO/ (accessed May 27, 2012).

CMS. 2012e. Pioneer ACO model. http://innovations.cms.go/initiatives/aco/pioneer (accessed May 27, 2012).

CMS. 2012f. Selected participants in the pioneer ACO model. http://innovations.cms.gov/Files/x/Pioneer-ACO-Model-Selectee-Descriptions-document.pdf (accessed May 27, 2012).

CMS. 2012g. Advance payment ACO model. http://innovations.cms.gov/initiatives/aco/advance-payment/ (accessed May 27, 2012).

CMS. 2012h. Fact sheets: First accountable care organizations under the Medicare shared savings program. http://www.cms.gov/apps/media/press/factsheet.asp?Counter=4334&intNumPerPage=10&checkDate=&checkKey=&srchType=1&numDays=3500&srchOpt=0&srchData=&keywordType=All&chkNewsType=6&intPage=&showAll=&pYear=&year=&desc=&cboOrder=date (accessed May 27, 2012).

CMS. 2012i. Potential future measures for hospital value-based purchasing program. http://www.hospitalcompare.hhs.gov/staticpages/for-consumers/value-based-purchasing.aspx (accessed May 27, 2012).

Coleman, E. A., C. Parry, S. Chalmers, and S. J. Min. 2006. The care transitions intervention: Results of a randomized controlled trial. Archives of Internal Medicine 166(17):1822-1828.

Croasdale, M. 2008. Jinx of the J-1 visas: IMGs finding other paths to residency, http://www.ama-assn.org/amednews/2008/03/10/prsa0310.htm (accessed May 23, 2012).

Crouse, B. J., and R. L. Munson 2006. The effect of the J-1 visa waiver on rural Wisconsin. Wisconsin Medical Journal 105(7):16-20.

CRS (Congressional Research Service). 2010. Medicare hospital readmissions: Issues, policy options, and PPACA. CRS Report for Congress, 7-7500. http://www.ncsl.org/documents/health/Medicare_Hospital_Readmissions_and_PPACA.pdf (accessed August 15, 2012).

Ebell, M. H. 2008. Research letter: Future salary and US residency fill rate revisited. Journal of the American Medical Association 300(10):1 131-1 132.

Escarce, J. J., D. Polsky, G. D. Wozniak, M. V. Pauly, and P. R. Kletke. 1998. Health maintenance organization location and the practice location decisions of new physicians. Medical Care 36(1 1):1555-1566.

Fairman, J., J. Rowe, S. Hassmiller, and D. Shalala. 2011. Broadening the scope of nursing practice. New England Journal of Medicine 364(3):193-196.

Fiegl, C. 2012. American Medical News: More physicians calling the shots in latest round of ACOs. http://www.ama-assn.org/amednews/m/2012/04/23/gl10423.htm (accessed May 27, 2012).

Fink, K. S., R. L. Phillips, G. E. Fryer, and N. Koehn. 2003. International medical graduates and the primary care workforce for rural underserved areas. Health Affairs 22(2):255-262.

Frank, R. G. 1985. Pricing and location of physician services in mental health. Inquiry 38:1 15-133.

Friedberg, M. W., P. S. Hussey, and E. C. Schneider. 2010. Primary care: A critical review of the evidence on quality and costs of health care. Health Affairs 29(5):766-772.

Gagne, R., and P. T. Leger. 2005. Determinants of physicians’ decisions to specialize. Health Economics 14:721-735.

GAO (Government Accountability Office). 1995. Opportunities to stretch scarce dollars and improve provider placement. GAO HEHS-996-28. Washington, DC: GAO. http://www.gao.gov/products/HEHS-96-28 (accessed May 23, 2012).

GAO. 2001. Health workforce: Ensuring adequate supply and distribution remains challenging. GAO-01-1042T. Washington, DC: GAO. http://gao.justia.com/department-of-health-and-human-services/2001/8/ health-workforce-gao-01-1042t/GAO-01-1042T-full-report.pdf (accessed August 9, 2012).

Gaskin, D., and C. Hoffman. 2000. Racial and ethnic differences in preventable hospitalizations across 10 states. Medical Care Research and Review 57(1):85-107.

Ginsburg, P. 2011. Spending to save—ACOs and the Medicare shared savings program. New England Journal of Medicine 364:2085-2086.

Glickman, S. W., F. S. Ou, E. R. DeLong, M. T. Roe, B. L. Lytle, J. Mulgund, J. S. Rumsfeld, W. B. Gibler, E. M. Ohman, K. A. Schulman, and E. D. Peterson. 2007. Pay for performance, quality of care, and outcomes in acute myocardial infarction. Journal of the American Medical Association 297(21):2373-2380.

Gold, J. 2012. FAQ on ACOs: Accountable care organizations, explained. Kaiser Health News. http://www.kaiserhealthnews.org/stories/2011/january/13/aco-accountable-care-organization-faq.aspx (accessed August 15, 2012).

Goodman, D. C., and K. Grumbach. 2008. Does having more physicians lead to better health system performance? Journal of the American Medical Association 299(3):335-337.

Grossbart, S. R. 2006. What’s the return? Assessing the effect of “pay-for-performance” initiatives on the quality of care delivery. Medical Care Research and Review 63(1):29S-48S.

Guterman, S., K. Davis, S. Schoenbaum, and A. Shih. 2009. Using Medicare payment policy to transform the health system: A framework for improving performance. Health Affairs 28(2):238-250.

Hagopian, A., K. Johnson, M. Fordyce, S. Blades, and G. L. Hart. 2003. The role of the rural hospital flexibility program in addressing workforce issues. Prepared by WWAMI Rural Health Research Center for Health Resources and Services Administration’s Office of Rural Health Policy. Seattle, WA: University of Washington.

Hancock, C., A. Steinbach, T. S. Nesbitt, S. R. Adler, and C. L. Auerswald. 2009. Why doctors choose small towns: A developmental model of rural physician recruitment and retention. Social Science & Medicine 69(9):1368-1376.

Hannan, E. L., M. J. Racz, G. Walford, T. J. Ryan, W. Isom, E. Bennett, and R. H. Jones. 2003. Predictors of readmission for complications of coronary artery bypass graft surgery. Journal of the American Medical Association 290(6):773-780.

Hansen, L. O., R. S. Young, K. Hinami, A. Leung, and M. V. Williams. 2011. Interventions to reduce 30-day rehospitalization: A systematic review. Annals of Internal Medicine 155:520-528.

Hart, L. G., M. J. Pirani, and R. Rosenblatt. 1994. Most rural towns lost physicians after their hospital closed. Rural Development Perspectives 10(1):17-21.

HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems Survey). 2012. HCAHPS fact sheet. http://www.hcahpsonline.org/files/HCAHPS%20Fact%20Sheet%20May%202012.pdf (accessed May 18, 2012).

HHS (Department of Health and Human Services). 2011a. Report to Congress: National strategy for quality improvement in health care. http://www.healthcare.gov/center/reports (accessed August 10, 2011).

HHS. 2011b. HHS launches new Affordable Care Act initiative to strengthen primary care. http://www.hhs.gov/news/press/2011pres/09/20110928a.html (accessed May 27, 2012).

Hicks, L. S., A. J. O’Malley, T. A. Lieu, T. Keegan, N. L. Cook, B. J. McNeil, B. E. Landon, and E. Guadagnoli. 2006. The quality of chronic disease care in U.S. community health centers. Health Affairs 25(6):1712-1723.

HRSA (Health Resources and Services Administration). 2012a. Shortage designations: Health professional shortage areas and medically underserved areas/populations. http://bhpr.hrsa.gov/shortage/ (accessed February 12, 2012).

HRSA. 2012b. National Advisory Council (NAC) on the National Health Service Corps. Executive summary January 19-20, 2012. http://nhsc.hrsa.gov/corpsexperience/aboutus/nationaladvisorycouncil/meetingsummaries/index.html (accessed May 22, 2012).

HRSA. 2012c. Area Health Education Centers (AHECs). http://bhpr.hrsa.gov/grants/areahealtheducationcenters/ta/Trainings/veterans/vmhtrainedahecs.pdf (accessed May 23, 2012).

HRSA. 2012d. Health centers: Where to go for care you can afford. http://www.hrsa.gov/ourstories/healthcenter/healthcenterweek.html.

IOM (Institute of Medicine). 2010. The future of nursing: Leading change, advancing health. Washington, DC: The National Academies Press.

Jencks, S. F. 2010. Defragmenting care (editorial). Annals of Internal Medicine 153:757-758.

Jencks, S. F., M. V. Williams, and E. A. Coleman. 2009. Rehospitalizations among patients in the Medicare fee-for-service program. New England Journal of Medicine 360(14):1418-1428.

Jha, A. K., K. E. Joynt, E. J. Orav, and A. M. Epstein. 2012. The long term effects of premier pay for performance on patient outcomes. New England Journal of Medicine 366(17):1606-1615.

Kahn, T. R., A. Hagopian, and K. Johnson. 2010. Retention of J-1 Visa Waiver Program physicians in Washington state’s Health Professional Shortage Areas. Academic Medicine 85(4):614-621.

Landon, B. E., S. T. Normand, A. Lessler, J. O’Malley, S. Schmaltz, J. M. Loeb, and B. J. McNeil. 2006. Quality of care for the treatment of acute medical conditions in U.S. hospitals. Archives of Internal Medicine 166(22):2511-2517.

Lindenauer, P. K., D. Remus, S. Roman, M. B. Rothberg, E. M. Benjamin, and D. W. Bratzler. 2007. Public reporting and pay for performance in hospital quality improvement. New England Journal of Medicine 356(5):486-496.

MedPAC (Medicare Payment Advisory Commission). 2007. Payment policy for inpatient readmissions: Report to the Congress: Reforming the delivery system. Washington, DC: MedPAC.

MedPAC. 2011. Report to the Congress: Medicare payment policy. Washington, DC: MedPAC.

Mertz, E., R. Jain, J. Breckler, E. Chen, and K. Grumbach. 2007. Foreign v. domestic education of physicians for the United States: A case study of physicians of south Asian ethnicity in California. Journal of Health Care for the Poor and Underserved 18(4):984-993.

Mick, S. S., S. Y. D. Lee, and W. P. Wodchis. 2000. Variations in geographical distribution of foreign and domestically trained physicians in the United States: “Safety nets” or “surplus exacerbation”? Social Science and Medicine 50(2):185-202.

Naylor, M. D., and E.T. Kurtzman. 2010. The role of nurse practitioners in reinventing primary care. Health Affairs 29(5):893-899.

Naylor, M. D., L. H. Aiken, E. T. Kurtzman, D. M. Olds, and K. B. Hirschman. 2011. The care span: The importance of transitional care in achieving health reform. Health Affairs 30(4):746-754.

Newhouse, J., A. P. Williams, W. Bennett, and W. B. Schwartz. 1982. Where have all the doctors gone? Journal of the American Medical Association 247(17):2392-2396.

Newhouse, R. P., J. Stanik-Hutt, K. M. White, M. Johantgen, E. B. Bass, G. Zangaro, R. F. Wilson, L. Fountain, D. M. Steinwachs, L. Heindel, and J. P. Weiner. 2011. Advanced practice nurse outcomes 1990-2008: A systematic review. Nursing Economic$ 29(5):230-250.

Nicholson, S. 2008. Medical career choices and rates of return. In Incentives and choice in health care, edited by F. A. Sloan and H. Kasper. Cambridge, MA: MIT Press.

OIG (Office of Inspector General, Department of Health and Human Services). 1994. National Health Service Corps: A survey of providers, facilities, and staff. OEI-09-91-01310. Washington, DC: OIG.

OIG. 1995. Area health education centers: A role in enhancing the rural practice environment. OEI-01-93-00570. Washington, DC: OIG.

Pathman, D. E., T. R. Konrad, and T. C. Ricketts. 1992. The comparative retention of National Health Service Corps and other rural physicians—results of a 9-year follow-up study. Journal of the American Medical Association 268(12):1552-1558.

Pathman, D. E., T. R. Konrad, and T.C. Ricketts. 1994a. The National Health Service Corps experience for rural physicians in the late 1980s. Journal of the American Medical Association 272(17):1341-1348.

Pathman, D. E., T. R. Konrad, and T. C. Ricketts. 1994b. Medical education and the retention of rural physicians. Health Services Research 29(1):39-58.

Pathman, D. E., D. H. Taylor, T. R. Konrad, T. S. King, T. Harris, T. M. Henderson, J. D. Bernstein, T. Tucker, K. D. Crook, C. K. Spaulding, and G. G. Koch. 2000. State scholarship, loan forgiveness, and related programs—the unheralded safety net. Journal of the American Medical Association 284(16):2084-2092.

Pathman, D. E., T. R. Konrad, T. S. King, D. H. Taylor, and G. G. Koch. 2004. Outcomes of states’ scholarship, loan repayment, and related programs for physicians. Medical Care 42(6):560-568.

Pathman, D. E., T. C. Ricketts, and T. R. Konrad. 2006a. How adults’ access to outpatient physician services relates to the local supply of primary care physicians in the rural southeast. Health Services Research 41(1):79-102.

Pathman, D. E., G. E. Fryer, Jr., R. L. Phillips, J. Smucny, T. Miyoshi, and L. A. Green. 2006b. National Health Service Corps staffing and the growth of the local rural non-NHSC primary care physician workforce. Journal of Rural Health 22(4):285-293.

Phillips, R. L., Jr., M. S. Dodoo, S. Petterson, I. Xierali, A. Bazemore, B. Teevan, K. Bennett, C. Legagneur, and J. Rudd. 2009. Specialty and geographic distribution of the physician workforce: What influences medical student & resident choices? Washington, DC: American Academy of Family Practice.

Pohl, J. M., C. Hanson, J. Newland, and L. Cronenwett. 2010. Unleashing the full potential of nurse practitioners to deliver primary care and lead health care teams. Health Affairs 29(5):900-905.

Politzer, R. M., L. Q. Trible, T. D. Robinson, D. Heard, D. L. Weaver, S. M. Reig, and M. Gaston. 2000. The National Health Service Corps for the 21st century. Journal of Ambulatory Care Management 23(3):70-85.

Polsky, D., P. R. Kletke, G. D. Wozniak, and J. J. Escarce. 2000. HMO penetration and the mobility of practicing physicians. Journal of Health Economics 19(5):793-809.

Porterfield, D. S., T. R. Konrad, C. Q. Porter, K. Leysieffer, R. M. Martinez, R. Niska, and F. Potter. 2003. Caring for the underserved: Current practice of alumni of the National Health Service Corps. Journal of Health Care for the Poor and Underserved 14(2):256-271.

Probst, J. C., M. E. Samuels, T. V. Shaw, G. L. Hart, and C. Daly. 2003. The National Health Service Corps and Medicaid inpatient care: Experience in a southern state. Southern Medical Journal 96(8):775-783.

Quinn, K. J., K. Y. Kane, J. J. Stevermer, W. D. Webb, J. L. Porter, H. A. Williamson, Jr., and M. C. Hosokawa. 2011. Influencing residency choice and practice location through a longitudinal rural pipeline program. Academic Medicine 86(11).

Rabinowitz, H. K., M. P. Rosenthal, J. J. Diamond, and T. N. Turner. 1993. Alternate career choices of medical students: Their relationship to choice of specialty. Family Medicine 25(10):665-667.

Ricketts, T. C., and R. Randolph. 2007. Urban-rural flows of physicians. Journal of Rural Health 23(4):277-285.

Ricketts, T. C., and R. Randolph. 2008. The diffusion of physicians. Health Affairs 27(5):1409-1415.

Rooks, L., R. T. Watson, and J. O. Harris. 2001. A primary care preceptorship for first-year medical students coordinated by an Area Health Education Center program: A six-year review. Academic Medicine 76(5):489-492.

Rosenblatt, R. A. 2010. Commentary: Do medical schools have a responsibility to train physicians to meet the needs of the public? The case of persistent rural physician shortages. Academic Medicine 85(4):572-574.

Rosenblatt, R. A., C. H. A. Andrilla, T. Curtin, and L. G. Hart. 2006. Shortages of medical personnel at community health centers: Implications for planned expansion. Journal of the American Medical Association 295:1042-1049.

Rosenthal, M. B., A. Zaslavsky, and J. P. Newhouse. 2005. The geographic distribution of physicians revisited. Health Services Research 40(6 Pt 1):1931-1952.

Ryan, A. M. 2009. Effects of the Premier hospital quality incentive demonstration on Medicare patient mortality annd cost. Health Services Research 44(3):821-842.

Ryan, A. M., J. Blustein, and L. P. Casalino. 2012. Medicare’s flagship test of pay-for-performance did not spur more rapid quality improvement among low-performing hospitals. Health Affairs 31(4):797-805.

Salinsky, E. 2010. Health care shortage designations: HPSA, MUA, and TBD. National Health Policy Forum Background Paper No. 75. Washington, DC: The George Washington University. http://nhpf.org/library/details.cfm/2789 (accessed May 22, 2012).

Shugarman, L. R., and D. O. Farley. 2003. Shortcomings in Medicare bonus payments for physicians in underserved areas. Health Affairs 22(4):173-178.

Sloan, F. A. 1970. Lifetime earnings and physicians’ choice of specialty. Industrial and Labor Relations Review 24:47-56.

UDS (Uniform Data Systems). 2012. Health center data, 2010. http://bphc.hrsa.gov/healthcenterdatastatistics/index.html (accessed February 15, 2012).

Werner, R. M., J. T. Kolstad, E. Z. Stuart, and D. Polsky. 2011. The effect of pay-for-performance in hospitals: Lessons for quality improvement. Health Affairs 30(4):690-698.

The White House. 2012. The Obama administration and community health centers. http://www.whitehouse.gov/sites/default/files/05-01-12_community_health_center_report.pdf (accessed May 25, 2012).

Wilson, N., I. Cooper, E. D. Vries, S. Reid, T. Fish, and B. Marais. 2009. A critical review of interventions to redress the inequitable distribution of healthcare professionals to rural and remote areas. Rural and Remote Health 9:1060.