Changing Demographics: Implications for Physicians, Nurses, and Other Health Workers
Printer-friendly Changing Demographics (Acrobat/pdf)

Spring 2003

U.S. Department of Health and Human Services
Health Resources and Services Administration
Bureau of Health Professions
National Center for Health Workforce Analysis

bhpr.hrsa.gov/healthworkforce/

 

TABLE OF CONTENTS

	COVER
	EXECUTIVE SUMMARY
	INTRODUCTION AGING OF THE POPULATION POPULATION FORECASTS IMPLICATIONS OF AN AGING POPULATION FOR THE DEMAND FOR HEALTH WORKER Increasing Demand for Health Care Services Increasing Demand for Health Workers IMPLICATIONS OF AN AGING POPULATION FOR THE SUPPLY OF HEALTH WORKERS Physician Supply Nurse Supply IMPLICATIONS OF AN AGING POPULATION FOR THE ECONOMICS OF THE HEALTH CARE SYSTEM CHANGING RACIAL AND ETHNIC COMPOSITION OF THE POPULATION POPULATION FORECASTS IMPLICATIONS OF THE CHANGING RACIAL AND ETHNIC COMPOSITION OF THE POPULATION FOR THE DEMAND FOR HEALTH WORKERS IMPLICATIONS OF THE CHANGING RACIAL AND ETHNIC COMPOSITION OF THE POPULATION FOR THE SUPPLY OF HEALTH WORKERS Physician Supply Nurse Supply GEOGRAPHIC LOCATION OF THE POPULATION POPULATION PROJECTIONS AND REGIONAL GROWTH PATTERNS EVOLVING TRENDS IN URBANIZATION URBAN DEMOGRAPHY AND THE EFFECTS ON PHYSICIAN LOCATIONS MODELING THE IMPACT OF CHANGING DEMOGRAPHICS ON THE FUTURE DEMAND FOR HEALTH PROFESSIONALS PHYSICIAN AGGREGATE REQUIREMENTS MODEL Modeling Physician Requirements Modeling Requirements for Physical Therapists, Optometrists, and Podiatrists NURSING DEMAND MODEL SUMMARY AND CONCLUSIONS
	REFERENCES

 

	EXECUTIVE SUMMARY
	The size and characteristics of the future health workforce are determined by the complex interaction of the health care operating environment, economic factors, technology, regulatory and legislative actions, epidemiological factors, the health care education system and demographics. Efforts over the past several decades to model the supply of and demand for health workers show there is a lack of consensus on the relationship between the health workforce and its determinants, the future values of many of these determinants, and forecasters' assumptions. The Workforce Analysis Branch of the Bureau of Health Professions (BHPr), Health Resources and Services Administration (HRSA), commissioned a report synthesizing the literature on one set of factors that will have a profound impact on the future health workforce-changing demographics-and discussing its implications for the health workforce. In addition, BHPr commissioned the update of two requirements forecasting models: the Physician Aggregate Requirements Model (PARM) and the Nursing Demand Model (NDM). The major findings of the literature and these two demand models are the following.
	Population Aging
	If health care consumption patterns and physician productivity remained constant over time, the aging population would increase the demand for physicians per thousand population from 2.8 in 2000 to 3.1 in 2020. Demand for full-time-equivalent (FTE) registered nurses per thousand population would increase from 7 to 7.5 during this same period. In 2000, physicians spent an estimated 32 percent of patient care hours providing services to the age 65 and older population. If current consumption patterns continue, this percentage could increase to 39 percent by 2020. The aging of the health workforce raises concerns that many health professionals will retire about the same time that demand for their services is increasing. Furthermore, the declining proportion of the population age 18 to 30 raises concerns regarding the ability to attract a sufficient number of new health workers. The rise in health care expenditures associated with the rapid increase in the elderly population will likely place additional pressures on the Medicaid and Medicare programs, as well as private insurers, to control health care costs. Such measures would likely decrease the demand for and supply of health professionals. The aging population could result in rising average patient acuity, which could in turn require higher nurse and physician staffing levels. One countervailing trend is that tomorrow's elderly might have lower disability rates than today's elderly, controlling for age, because of improvements in economic resources, education levels, lifestyle, public health, and medical technology.
	Increasing Racial and Ethnic Diversity
	The literature suggests that Hispanics and non-whites have different patterns of health care use compared to non-Hispanic whites. Disparities in access to care account for part of the difference in utilization. Demand for health care services by minorities is increasing as minorities grow as a percentage of the population. Between 2000 and 2020, the percentage of total patient care hours physicians spend with minority patients will rise from approximately 31 percent to 40 percent.
	Minorities are underrepresented in the physician and nurse workforce relative to their proportion of the total population. As minorities constitute a larger portion of the population entering the workforce, their representation in the physician and nurse professions will increase. The U.S. will increasingly rely on minority caregivers. Minority physicians have a greater propensity than do non-minority physicians to practice in urban communities designated as physician shortage areas. An increase in minority representation in the physician workforce could improve access to care for the population in some underserved areas.
	Geographic Location of the Population
	Geographic variation in population growth rates and in determinants of health worker demand and supply highlight the importance of developing forecasting models that can make State-level and sub-State level forecasts. Although an increasing proportion of the U.S. population resides in urban areas, a substantial proportion of the population will continue to reside in rural areas. Many of these rural areas are currently designated as physician shortage areas. Pockets of urban areas will continue to have a high concentration of minorities. Many of these areas are currently designated as physician shortage areas. Efforts to increase the supply of health professionals in these areas must deal with economic, cultural and language considerations.
	Forecasting the Impact of Changing Demographics and Other Factors on Physician Requirements
	The PARM forecasts requirements for allopathic (MD) and osteopathic (DO) physicians providing patient care in 19 specialties as well as physicians in non-patient-care activities. Requirements are demand-based and rely on current and forecasted patterns of health care use, physician staffing patterns, and medical insurance prevalence rates. We consider forecasts under five scenarios (Exhibit ES.1). Scenario 1, Status Quo, forecasts physician requirements under the assumption that patterns of health care use, medical insurance coverage, and physician productivity remain constant over time. Under this scenario, total requirements for physicians would increase from approximately 781,300 in 2000 to 1,038,200 in 2020 (a 33 percent increase). Scenario 2, Baseline, is our best estimate of demand for physicians based on changing demographics and projected trends in the other factors (e.g., insurance coverage and economic considerations). Under this scenario, physician requirements would increase to 996,400 in 2020 (a 28 percent increase). Scenario 3, Universal Coverage, assumes that the entire U.S. population has medical insurance. Under this scenario, the uninsured population is placed into the insured fee-for-service and health maintenance organization (HMO) settings based on the current proportion of the insured population in each of those two settings. Under this scenario, total demand for physicians would have been an estimated 817,600 in 2000, increasing to an estimated 1,092,400 (a 40 percent increase over the 2000 baseline level). Scenario 4 is universal health care coverage with 100 percent of the population enrolled in a health maintenance organization. Under this scenario, total requirements would have been an estimated 781,900 in 2000, increasing to an estimated 1,059,900 in 2020 (a 36 percent increase over the 2000 baseline level).
	Scenario 5, Non-minority Rates, assumes that minorities have rates of medical insurance coverage similar to non-Hispanic whites within each demographic group defined by age and sex. Under this scenario, demand for physicians would have been an estimated 802,400 in 2000, increasing to an estimated 1,072,000 in 2020 (a 37 percent increase over the 2000 baseline level).

	Exhibit ES.1 Forecasted Physician Requirements
	Scenario 2000 2020 1: Status Quo 781,282 1,038,234 2: Baseline 781,282 996,387 3: Universal Coverage 817,615 1,092,381 4: 100 percent HMO 781,889 1,059,907 5: Non-minority Rates 802,356 1,072,048

	The PARM also forecasts requirements for three non-physician specialties: physical therapy, podiatry, and optometry. Based on available data and studies, the requirements for all three professions are projected to increase, between 2000 and 2020, at rates equal to or slightly greater than the growth in population.
	Forecasting the Impact of Changing Demographics and Other Factors on Nurse Requirements
	The NDM forecasts demand-based requirements for FTE registered nurses (RNs), licensed practical nurses (LPNs), nurse aides and home health aides (NAs). Although the NDM forecasts requirements at the State level, in this report we present only national-level forecasts (Exhibit ES.2). Under a baseline scenario, which represents the forecasts most likely to occur based on changing demographic and projected trends in other determinants of nurse demand, total requirements for FTE RNs would increase from approximately 2 million in 2000 to 2.8 million in 2020 (a 41 percent increase). Requirements for FTE LPNs would increase from 618,000 in 2000 to 905,000 in 2020 (a 46 percent increase). There would also be an increase in FTE nurse aide and home health aide requirements from 1.5 million in 2000 to 2.3 million in 2020 (a 50 percent increase). Demand for nurses and nurse aides will continue to grow in hospitals during the next two decades, but at a slower rate than for the nursing professions as a whole. The exception results from strong growth in demand for RNs in hospital outpatient settings as technological innovations and managed care trends shift patients from inpatient to outpatient care. The fastest growth in demand will occur in nursing facilities and home health. Under a status quo scenario where patterns of per capita health care use and nurse staffing remain constant over time, the requirement for nurses and nurse aids increases at a slower rate than under the baseline scenario.

	Exhibit ES.2 Forecasted FTE Nurse Requirements
	Baseline Scenario Status Quo Scenario   2000 2020 2020 Registered nurses 2,001,198 2,822,388 2,505,747 Licensed practical nurses 617,946 905,159 787,329 Nurse aides and home health aides 1,545,722 2,323,518 1,983,582

	Findings from the PARM and NDM, as well as the literature review, provide important insights on the impact of changing demographics on the health workforce. This report also identifies areas for additional research such as (a) factors changing the per capita use of health care services, (b) the paucity of information on the relationship between race/ethnicity and the supply of health workers, and (c) the need for models that can forecast demand for and supply of health workers at smaller geographic units of aggregation (e.g., at the sub-State level).

	INTRODUCTION
	The size and characteristics of the future health workforce are determined by the complex interaction of the health care operating environment, economic factors, technology, regulatory and legislative actions, epidemiological factors, the health care education system and demographics. Efforts over the past several decades to model the supply of and demand (or "requirements") for health workers show there is a lack of consensus on the relationship between the health workforce and its determinants, the future values of many of these determinants, and forecasters' assumptions. [1] See, for example, recent articles by Snyderman, Sheldon and Bischoff (2002), Weiner (2002), Grumbach (2002) and Reinhardt (2002) commenting on recent physician workforce projections by Cooper et al. (2002). Prescott (2000) discusses the lack of consensus as it pertains to modeling the nurse workforce. Furthermore, past forecasts of impending surpluses and shortages of health professionals often failed to materialize, leading to the general consensus that a much better understanding is needed about the dynamics affecting the supply of and demand for health professionals. The Workforce Analysis Branch of the Bureau of Health Professions (BHPr), Health Resources and Services Administration (HRSA), commissioned a report synthesizing the literature on one set of factors that will have a profound impact on the future health workforce-changing demographics. In addition, BHPr commissioned the updating of two requirements forecasting models: the Physician Aggregate Requirements Model (PARM) and the Nursing Demand Model (NDM). This report discusses findings from the literature review of the implications of important demographic trends for the health workforce. In addition, this report presents findings from the NDM and PARM to quantify the impact of changing demographics on demand for allopathic (MD) and osteopathic (DO) physicians, registered nurses (RNs), licensed practical nurses (LPNs), nurse aides and home health aides (NAs), physical therapists, optometrists, and podiatrists. This report also presents forecasts from the PARM and NDM for several scenarios with different assumptions regarding the future health care operating environment, the productivity of doctors and nurses, and other factors.
	Although the demographic trends discussed here have implications for the entire health workforce, the discussion in this report is heavily tilted towards the physician and nursing professions. Reasons for this focus include the dominance of these professions in the health workforce literature, the focus on these professions by government commissions and policy makers, and the availability of the PARM and NDM for forecasting requirements for physicians and nurses.
	Demographics are a major determinant of the size and characteristics of the future health workforce, and demographic trends can be extrapolated with reasonable accuracy one or two decades into the future. In addition to the growth in size of the U.S. population in future decades, three demographic trends have profound implications for the future health workforce: First, the population is aging and the size of the elderly population will increase substantially. An aging population will place greater demands on the health care system at the same time that many health professionals will be retiring. Also, as the population ages there will be a continuing shift in the type and setting of services provided. Second, the population is becoming more racially and ethnically diverse. Concerns that minorities are underrepresented in the health workforce have both equity implications for people who need health care services and efficiency implications for the health care system. As minorities constitute a larger proportion of persons entering the workforce, the U.S. population will increasingly rely on minority health workers for their care. Third, the population is shifting geographically and a significant portion of the U.S. population will continue to reside in areas with persistent shortages of health workers. These trends highlight the need for forecasting models that can make State-level and sub-State-level forecasts of health worker supply and demand.
	Other demographic trends with implications for the future supply of and demand for health workers include changes in fertility patterns, family size and composition, longevity, immigration, and overall health of the population. These trends are discussed within the context of the three major trends discussed above. In both the PARM and NDM, requirements are defined as the number of health professionals demanded based on the level of health care services that society is willing to purchase given population needs and economic considerations. Other authors have used “need” to define requirements, where need is based on the analyst’s assessment of what constitutes an adequate supply of health workers, independent of society's willingness or ability to purchase services. Using the PARM and NDM, we forecast future demand for health care services and the derived demand for 19 physician specialties, nurses, and the other health workers listed previously. We forecast a “status quo” scenario that assumes no change in per capita health care utilization patterns, health worker productivity, and health worker staffing patterns. Under such a scenario, between the years 2000 and 2020, changing demographics would cause an estimated 30 percent increase in inpatient days, a 20 percent increase in outpatient visits, and a 17 percent increase in emergency department visits at general, short-term hospitals.  Inpatient days at non-general and long-term hospitals would increase by an estimated 33 percent; the number of nursing facility residents would increase by 40 percent; the number of home health visits would increase by 36 percent; and the number of visits to physicians’ offices would increase by 23 percent. The change in demand for health care services would increase requirements for physicians by approximately 33 percent, although the increase in requirements would vary by medical specialty. For example, requirements for cardiologists would increase by an estimated 52 percent while requirements for pediatricians would increase by an estimated 11 percent. Requirements would increase approximately 28 percent for RNs, 30 percent for LPNs, and 33 percent for nurse aides (including home health aides).

	Although demographics are a dominant determinant of the demand for health workers, other important factors are the characteristics of the future health care system, economic considerations, technological advances, and population needs. A detailed discussion of these trends is outside the scope of this project; however, the extant literature in this area is relatively large. [2] The report: The Impact of the Restructuring of the U.S. Health Care System on the Physician Workforce and Vulnerable Populations (The Lewin Group, 1998), contains a literature review that discusses many of these trends. Using the PARM and NDM, we forecast future requirements for selected health care professions under alternative scenarios regarding the future health care operating environment. The baseline scenario in both the PARM and NDM produce the forecasts that are most likely to occur based on changing demographics and projected trends in the factors listed above (e.g., trends in insurance coverage and economic considerations). The baseline forecasts for physician requirements are slightly lower than under the status quo scenario (28 percent growth between 2000 and 2020 instead of 33 percent growth), and the change in requirements for individual physician specialties is quite different in some cases. Under the NDM’s baseline scenario, requirements for RNs grow faster than under the status quo scenario (41 percent growth between 2000 and 2020 instead of 28 percent growth), reflecting different assumptions about changes in average patient acuity levels and other factors. Under the baseline scenario, total requirements for LPNs, nurse aides, and home health aides rise faster than forecasts under the status quo scenario.

	The remaining sections in this report discuss the implications for the health workforce of the aging population (Section 2), the changing racial and ethnic composition of the population (Section 3), and population geographic location (Section 4). Each of these sections presents information on the demographic trend, discusses the implications of the trend on demand for health care services and derived demand for health workers, and discusses the implications for the supply of health workers. Section 5 describes the recently updated PARM and NDM and presents findings from these models. Section 6 summarizes the main findings of this effort and discusses areas for additional research.

	AGING OF THE POPULATION
	Increased longevity and the Major Findings: If health care consumption patterns and physician productivity remained constant over time, the aging population would increase the demand for physicians per thousand population from 2.8 in 2000 to 3.1 in 2020. Demand for full-time-equivalent RNs per thousand population would increase from 7 to 7.5 during this same period. In 2000, physicians spent an estimated 32 percent of patient care hours providing services to the age 65 and older population. If current consumption patterns continue, this percentage could increase to 39 percent by 2020. The aging of the health workforce raises concerns that many health professionals will retire about the same time that demand for their services is increasing. Also, the elderly population will grow at a faster rate than the working-age population. The rise in health care expenditures associated with the rapid increase in the elderly population will likely place pressures on the Medicaid and Medicare programs to control health care costs. Such measures would likely decrease the demand for and supply of health professionals. aging of the baby boom generation will contribute to a substantial increase in the size of the elderly population during the next few decades as well as the aging of the overall population. Four major implications of an aging population on the health workforce are the following. One, because the elderly have both greater and different health care needs than the non-elderly, the rapid growth in size of the elderly population could substantially increase overall demand for health care services and consequently the derived demand for health workers. Occupations and settings that disproportionately serve the elderly will experience the largest growth. If health care consumption patterns and physician productivity remained constant over time, the aging population would increase the demand for physicians per thousand population from 2.8 in 2000 to 3.1 in 2020. Demand for full-time-equivalent (FTE) RNs per thousand population would increase from 7 to 7.5 during this same period. Two, physicians will spend an increasing proportion of their time treating the elderly. Our analysis of multiple health care use databases suggests that in 2000 physicians spent an estimated 32 percent of total patient care hours providing services to the age 65 and older population. If current patterns continue, this percentage could increase to 39 percent by 2020. Three, the health workforce is aging along with the general population. As health professionals in the baby boom generation retire and as the pool of potential entrants to the health workforce (i.e., the population age 18 to 30) declines as a percentage of the total population, there is concern that the future supply of health professionals will be inadequate to meet demand. Four, the expected increase in health care expenditures attributed to the growing elderly population will likely place pressures on the Medicaid and Medicare programs to control health care costs. The ratio of working-to-retired Americans will likely decrease, placing budget pressures on other government programs that compete with funding for Medicaid and Medicare. Economic pressures to curb the growth in health care costs could result in policies to reduce the demand for and supply of health workers.

	2.1 Population Forecasts
	Census Bureau population projections show significant shifts in the age distribution (Exhibit 2.1) with the number of elderly increasing in absolute size and as a proportion of the total population (Exhibit 2.2). The number of elderly, defined as the "age 65 and over" population, will grow by over 50 percent between 2000 and 2020, and by an estimated 127 percent by 2050. Furthermore, the relative size of the elderly population is projected to increase from 12.6 percent of the population in 2000 to an estimated 16.5 percent in 2020. Between 2030 and 2050, one in five Americans will be elderly. The most rapidly growing demographic group among age categories is the "oldest elderly." This group is sometimes defined differently by researchers, but the most common definitions are the population age 75 and over, age 80 and over, and age 85 and over. [3] Two factors that contribute to researchers using different age breaks to define the oldest elderly are (1) differences in use of health care services, and (2) small sample size among the oldest elderly when using survey data. In 2000, there were approximately 16.6 million people age 75 and over, 9.2 million people age 80 and over, and 4.2 million people age 85 and over. By 2020, the number of people in these age groups could reach 22 million, 13 million, and 7 million, respectively.

	Exhibit 2.1. Age Distribution of U.S. Population
	Exhibit 2.1. Age Distribution of U.S. Population (Text Only)
	Age 2000 2020 2050 0-9 14.2% 13.5% 13.6% 10-19 14.5% 13.2% 13.5% 20-29 13.1% 13.3% 12.8% 30-39 15.2% 13.0% 12.4% 40-49 15.4% 11.6% 11.5% 50-59 11.1% 12.6% 11.0% 60-69 7.3% 11.8% 10.0% 70-79 5.9% 7.2% 7.6% 80-89 2.8% 2.9% 5.4% 90+ 0.6% 0.9% 2.3%
	Source: U. S. Census Bureau middle series population projections (Day, 1996).
	Exhibit 2.2. Projections of U.S. Elderly Population
	Year Mean Age Population 65+ (in millions) % of Population 65+ %increase from 2000 in 65+ population 2000 36.5 34.71 12.6 -- 2005 37.2 36.17 12.6     4.2 2010 37.8 39.41 13.2 13.5 2020 39.0 53.22 16.5 53.3 2030 39.9 69.38 20.0 99.9 2040 40.3 75.23 20.3 116.8 2050 40.3 78.86 20.0 127.2

	2.2 Implications of an Aging Population for the Demand for Health workers
	2.2.1 Increasing Demand for Health Care Services
	The greater medical needs of the elderly, combined with access to health care services through Medicare and Medicaid, have resulted in much higher per capita use of health care services for the elderly compared to the non-elderly. On a per capita basis, the elderly have more hospital inpatient days, outpatient visits, and emergency department visits. Relative to the non-elderly, they also have more home health visits per capita and are more likely to be in a long-term care facility. To illustrate these points, consider Exhibits 2.3 through 2.8 that contain estimates of per capita health care use by age, sex, and urban or rural location for six health care settings modeled in the NDM. The most profound differences in per capita utilization exist across age groups; however, there are also important differences in per capita utilization by sex and by urban or rural location. Many of the following estimates are for 1996, the base year in the NDM, although more recent data are available for some settings.
	An analysis of the 1996 Health Cost Utilization Project (HCUP) database finds that with the exception of the age 0-4 population, the number of inpatient days in general, short-term hospitals per 1,000 population increases substantially with age for both men and women, in both rural and urban areas (Exhibit 2.3). Analyses of other patient-level databases such as the National Hospital Ambulatory Medical Care Survey (NHAMCS), the National Home and Hospice Care Survey (NHHCS), and the National Nursing Home Survey (NNHS) produced estimates of per capita health care utilization in different settings for the eight age groups used in the NDM, by sex, and by urban or rural location. These are shown in Exhibits 2.4 through 2.8.

	Exhibit 2.3. Inpatient Days in General, Short-term Hospitals (per 1,000 population)
	Rural Urban Age Category Female Male Female Male 0-4 years 430 449 789 838 5-17 years 57 45 79 81 18-24 years 276 83 280 141 25-44 years 218 134 327 242 45-64 years 307 317 470 633 65-74 years 919 1,049 1,187 1,640 75-84 years 1,871 2,137 1,985 2,468 85 years and above 3,052 3,826 2,734 3,302
	Source: Analysis of the 1996 HCUP database with an adjustment so that rates applied to the population in 1996 equaled total inpatient days reported by the American Hospital Association (AHA). See Dall and Hogan (2002).

	Exhibit 2.4. Outpatient Visits in General, Short-term Hospitals (per 1,000 population)
	Rural Urban Age Category Female Male Female Male 0-4 years 1,472 2,967 985 3,519 5-17 years 783 1,838 651 1,548 18-24 years 954 3,418 592 876 25-44 years 931 2,472 485 1,290 45-64 years 1,464 2,818 833 1,793 65-74 years 2,365 2,593 2,671 2,152 75-84 years 4,841 1,933 4,033 1,896 85 years and above 5,081 1,709 5,734 1,685
	Source: Analysis of the 1996 NHAMCS database with an adjustment so that rates applied to the population in 1996 equaled total non-emergency, outpatient visits reported by the AHA. See Dall and Hogan (2002).

	Exhibit 2.5. Emergency Department Visits in General, Short-term Hospitals (per 1,000 population)
	Rural Urban Age Category Female Male Female Male 0-4 years 825 426 754 476 5-17 years 422 204 369 211 18-24 years 620 376 534 286 25-44 years 432 284 364 259 45-64 years 346 211 335 190 65-74 years 471 248 468 237 75-84 years 681 313 730 328 85 years and above 953 457 1,298 557
	Source: Analysis of the 1996 NHAMCS database with an adjustment so that rates applied to the population in 1996 equaled total emergency visits reported by the AHA. See Dall and Hogan (2002).

	Exhibit 2.6. Inpatient Days in Non-General and Long-term Hospitals (per 1,000 population)
	Rural Urban Age Category Female Male Female Male 0-4 years 0 0 24 33 5-17 years 0 1 17 25 18-24 years 2 2 27 56 25-44 years 4 4 64 85 45-64 years 23 19 169 198 65-74 years 131 145 411 514 75-84 years 221 284 695 664 85 years and above 234 201 773 806
	Source: Analysis of the 1996 HCUP database with an adjustment so that rates applied to the population in 1996 equaled total inpatient days reported by the AHA. See Dall and Hogan (2002).

	Exhibit 2.7. Home Health Visits (per 1,000 population)
	Rural Urban Age Category Female Male Female Male 0-17 years 420 400 427 406 18-44 years 232 169 403 190 45-64 years 1,497 1,367 1,180 702 65-74 years 8,032 5,230 5,332 3,570 75-84 years 22,211 13,327 12,607 9,485 85 years and above 33,507 29,117 17,534 13,429
	Source: Analysis of the 1995 NHHCS database with an adjustment so that rates applied to the population in 1998 equaled estimates of total home health visits paid for by Medicare, Medicaid and other sources in 1998. See Dall and Hogan (2002).

	Exhibit 2.8. Nursing Home Residents (Residents per 1,000 population)
	Urban & Rural Age Category Female Male 0-44 years 0.2 0.2 45-64 years 2.6 1.0 65-74 years 14.5 6.9 75-84 years 52.4 32.0 85 years and above 194.4 187.0
	Source: Analysis of the 1997 National Nursing Home Survey (NNHS). See Dall and Hogan (2002).

	Not only does per capita use of health care services within a delivery setting increase with age, but also the type of services used by the elderly (and the mix of health professionals who provide these services) differs from those of the non-elderly. To capture these differences in type of services received, the PARM uses physician-patient encounters in hospital inpatient and outpatient settings, in non-hospital office settings, and in other settings (e.g., nursing homes and home health) to forecast future demand for physician services by medical specialty. [4] The nature of a physician-patient encounter, as well as the length of the encounter, can vary substantially by medical specialty and delivery setting. Physician surveys, reported in the annual AMA publication Physician Socioeconomic Statistics, reveal that physicians typically spend more time per encounter with patients in hospital-based visits versus office visits that are not hospital-based. Encounters that involve surgical procedures often last two to five times longer, on average, than visits that do not involve surgical procedures. Consequently, the PARM forecasts demand for each physician specialty by health care setting, and the hospital inpatient setting is subdivided by whether or not a surgical procedure was performed. Even within a specialty, the types of services demanded might differ by age. For example, eye diseases such as cataracts and glaucoma are much more prevalent in the older population (White et al., 2000). Consequently, as the population ages, optometrists will likely see a shift in the type of services provided. An important question for modeling requirements for physicians and other health workers is whether these caregivers spend different amounts of time per encounter with the elderly relative to the non-elderly. Two databases used to update the PARM-the National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Care Survey (NHAMCS) Outpatient File-contain information on the amount of time physicians spent with patients during each encounter. To increase sample size, we combined the 1997, 1998, and 1999 NAMCS, and we combined the 1997, 1998, and 1999 NHAMCS. We tested the hypothesis that patient demographic characteristics and insurance status are determinants of the amount of time physicians spend per visit with patients in doctors' offices and hospital outpatient settings. We tested this hypothesis by estimating a series of regressions, using the ordinary least squares (OLS) criterion, with length of time as the dependent variable and dummy variables that indicate patient characteristics and insurance status as the exogenous variables. The dummy variables take on the value of 1 if the patient has that characteristic, and take on the value of 0 if the patient does not have that characteristic. We estimated separate regressions for each medical specialty. The regression results showed each of the exogenous variables (age, sex, race/ethnicity, and insurance status) to have a significant impact on the dependent variable (time per encounter) for some specialties but not for others. Even when statistically significant, the impact was in many cases quite small, less than two minutes per encounter. One caution when interpreting the regression results is that the R-squared statistic for every regression is extremely low, indicating that the exogenous variables in the model explain only a small proportion of the overall variation in length of time physicians spend with patients. The relatively large residual variance makes it more difficult to find a statistically significant relationship. Also, for some specialties the number of patients in a particular demographic group is small which reduces the precision of the estimates for those demographic groups. Exhibit 2.9 contains the regression results for encounters in doctors' offices, and Exhibit 2.10 contains the results for encounters in hospital outpatient settings. The column labeled AVG reports the average minutes per encounter for the reference group (non-Hispanic, white females age 55-64, insured in a fee-for-service arrangement). The other columns represent the marginal impact of the demographic characteristic or insurance status on minutes of physician time per encounter. Shaded boxes indicate marginal impacts, relative to the reference category, that are statistically different from zero at the 0.05 level of significance. To illustrate, consider the first specialty: general and family practitioners. The average time spent with the reference group is 18.36 minutes per encounter in doctors' offices (Exhibit 2.9). Time spent with men is just 6 seconds shorter than time spent with women, on average, after controlling for age, race/ethnicity, and insurance status. General and family practitioners spend, on average, 2.43 fewer minutes per encounter with patients age 0-17 and 1.08 fewer minutes per encounter with patients age 18-34 compared to the reference group of patients age 55-64. Both of these differences in average minutes per encounter are statistically different from zero at the 0.05 level of significance. General and family practitioners also spend 0.91 fewer minutes per encounter with African Americans and 0.53 fewer minutes per encounter with other minorities, relative to non-Hispanic whites, although only the estimate for African Americans is statistically different from zero. Time spent with patients in a health maintenance organization (HMO) is 0.81 minutes less than time spent with patients insured in a fee-for-service arrangement, while the time spent with uninsured patients is 0.74 minutes greater than that spent with patients covered under fee-for-service. Neither of these differences is large, however, and of the two, only the former is statistically different from zero. With respect to the other specialties shown in Exhibit 2.9, major regression effects noted are as follows: Sex. - Only orthopedic surgery and other surgical specialties show statistically significant differences for men and women. The time per encounter is in both cases greater for men than it is for women: an additional 0.66 minutes, on average, for orthopedic surgery, an additional 3.86 minutes for other surgical specialties. Age. - Of the sixteen specialties shown, ten display significant age effects with respect to at least one age group. General and family practitioners, for example, spend significantly fewer minutes per encounter with patients under 35; internal medicine (IM) subspecialists spend significantly fewer minutes per encounter with patients over 74; etc. Most of these effects, however, although statistically significant, are no more than a minute or two, with the following exceptions: physicians in other medical specialties spend over three minutes more per encounter with children under 18 while physicians in other surgical specialties spend almost seven minutes less per encounter with patients in that age group. Race/ethnicity. - Significant race/ethnicity effects are evident for ten of the specialties shown. African Americans spend significantly fewer minutes per encounter with physicians in four specialties (general and family practice, internal medicine subspecialties, cardiovascular disease, and other patient care) and significantly more minutes per encounter with ob/gyn's. Patients in the "other" minority category spend significantly fewer minutes per encounter with physicians in three specialties (general internal medicine, pediatrics, and psychiatry) and significantly more minutes per encounter with physicians in another three (other medical specialties, emergency medicine, and other patient care). The added 14.51 minutes per encounter for "other" minority patients seen by emergency medicine physicians is particularly noteworthy. Insurance status. - A marked insurance effect is also evident. HMO patients spend significantly fewer minutes per encounter with physicians in four specialties (general and family practice, pediatrics, orthopedic surgery, and other patient care) and significantly more minutes per encounter with physicians in four other specialties (IM subspecialties, cardiovascular disease, other surgical specialties, and psychiatry). Of these differences, only those for other surgical specialties (plus 3.82 minutes) and other patient care (minus 2.61) exceed 2 minutes. Somewhat surprisingly, there are no specialties for which uninsured patients receive fewer minutes per encounter, on average, than the reference group, whereas there are six specialties for which they receive more minutes on average. Those six are pediatrics, other medical specialties, general surgery, ophthalmology, other surgical specialties, and psychiatry. The added time per encounter, on average, is particularly great for physicians in other surgical specialties (an additional 11.44 minutes) and psychiatry (an additional 7.95). In addition to these observations, applicable to encounters in doctors' offices, observations of a similar nature are noted with respect to time spent in hospital outpatient clinics (Exhibit 2.10). General and family practitioners are seen to spend 24.06 minutes per encounter, on average, with members of the reference group. They spend slightly less time per encounter with men, less time with younger patients, more time with African Americans, less time with patients in the "other" minority category, more time with patients in HMOs, and less time with the uninsured. None of these differences, however, is statistically different from zero at the 0.05 level of significance.

	Exhibit 2.9. Minutes of Physician Time Spent with Patients in Doctors' Offices (by Patient Characteristics and Insurance Status)
	PARM Classification   Avg. Marginal Impact (deviation from average minutes) Sex Age Race/Ethnicity Insurance Male Female 0-17 18-34 35-54 55-64 65-74 75+ White AA Other FFS HMO None General Primary Care   Reference Category   Reference Category   Reference Category   Reference Category   General & family practice 18.36 -0.10 -2.43 -1.08 -0.43 -0.54 0.14 -0.91 -0.53 -0.81 0.74 General internal medicine 20.18 -0.29 -0.62 -0.02 0.31 0.16 0.55 -0.52 -1.48 -0.54 -0.34 Pediatrics 15.97 -0.23 a a a a a 0.25 -1.01 -0.66 3.31 Medical Specialties           IM subspecialties 23.77 -0.81 -3.03 -1.24 1.43 -0.51 -2.13 -2.58 -1.61 1.89 -3.60 Cardiovascular diseases 19.86 -0.15 3.28 0.76 0.73 1.85 0.94 -1.69 -1.08 1.60 -0.22 Other medical specialties 15.68 -0.08 3.04 1.08 1.69 -0.71 -0.18 0.46 1.51 0.64 2.01 Surgery           General surgery 18.05 0.17 -1.92 -1.35 -0.25 -0.71 -0.79 -0.00 1.17 -0.23 3.36 Obstetrics & gynecology 18.10 b -0.57 -2.21 0.12 0.10 0.91 0.97 0.54 -0.26 0.79 Otolaryngology 17.21 -0.05 -1.76 0.03 -0.13 -0.08 1.15 0.73 -0.32 -0.49 -1.14 Orthopedic surgery 15.77 0.66 0.14 0.96 0.27 -0.26 0.34 0.86 0.38 -0.83 1.07 Urology 16.80 0.42 0.89 2.13 1.34 0.06 -0.32 -0.90 -0.13 -0.43 -1.20 Ophthalmology 17.11 0.33 1.56 0.08 0.11 0.10 -1.06 0.92 0.54 -0.03 2.77 Other surgical specialties 15.15 3.86 -6.54 -4.37 0.09 -3.73 -0.18 0.41 0.77 3.82 11.44 Other Patient Care         Emergency medicine 7.15 2.83 -1.80 2.28 3.77 3.04 1.69 21.51 14.51 4.74 -2.61 Psychiatry 34.14 0.70 0.18 0.13 0.77 2.01 -3.17 -0.04 -2.70 1.72 7.95 Other specialties 25.31 -0.23 0.08 1.33 1.08 1.93 0.39 -3.38 2.51 -2.61 -0.52
	Source: Analysis of the 1997, 1998, and 1999 NAMCS. Note: Shaded boxes indicate marginal impacts, relative to the reference category, that are statistically different from zero at the 0.05 level of significance. a The large majority of patients seen by pediatricians are age 17 and younger, so the sample size of adults seen by pediatricians is insufficient to obtain reliable estimates by age group. b This physician specialty saw no patients with this characteristic.
	Exhibit 2.10. Minutes of Physician Time Spent with Patients in Hospital Outpatient Clinics (by Patient Characteristics and Insurance Status)
	PARM Classification   Avg. Marginal Impact (deviation from average minutes) Sex Age Race/Ethnicity Insurance Male Female 0-17 18-34 35-54 55-64 65-74 75+ White AA Other FFS HMO None General Primary Care   Reference Category   Reference Category   Reference Category   Reference Category   General & family practice 24.06 -0.33 a -1.96 -0.54 0.80 1.21 1.01 -0.82 0.94 -0.98 General internal medicine 26.23 -1.25 a -1.48 -1.98 -2.56 -0.53 -0.00 -2.64 4.43 -1.53 Pediatrics 25.83 -1.14 a a a a a -0.89 -1.89 0.12 2.28 Medical Specialties           IM subspecialties 26.85 -0.38 a -0.34 -0.15 1.22 -0.43 -1.36 -2.90 1.32 -1.81 Cardiovascular diseases 25.88 0.66 -2.06 4.18 0.95 2.87 0.31 -1.16 -1.98 -2.82 -4.47 Other medical specialties 25.28 -1.36 -3.98 -3.25 -1.28 1.75 2.50 1.76 0.013 1.06 1.21 Surgery           General surgery 22.97 0.59 -2.50 -1.46 0.65 0.64 -2.80 -1.10 1.34 -0.97 -1.27 Obstetrics & gynecology 21.76 b -2.43 -3.04 -0.04 2.36 -1.24 0.98 -0.38 -0.11 0.81 Otolaryngology 17.10 1.00 -1.64 -1.12 1.64 1.89 4.23 2.38 0.20 0.69 -0.27 Orthopedic surgery 27.01 -2.36 b -3.90 1.27 2.98 -1.56 0.31 -0.34 -2.17 -4.96 Urology 22.54 2.33 -3.73 -4.92 -4.26 -0.72 0.92 0.94 1.30 -1.04 1.01 Ophthalmology 24.21 0.48 -2.38 1.47 -0.70 4.48 7.05 2.25 -3.13 -4.15 -5.79 Other surgical specialties 24.67 -0.28 1.34 -2.51 -1.48 1.61 3.53 0.18 -1.58 3.49 1.51 Other Patient Care           Psychiatry 27.06 -0.04 2.68 2.05 0.07 1.37 4.08 1.47 -2.04 2.77 -0.78 Other specialties 23.32 0.80 b -2.44 -1.12 -3.12 -2.21 3.65 -6.32 -2.99 -7.19
	Source: Analysis of the 1997, 1998, and 1999 NHAMCS. Note: Shaded boxes indicate marginal impacts, relative to the reference category, that are statistically different from zero at the 0.05 level of significance. a The specialty imputation method identified the physician of patients age 0-17 with general primary care diagnoses or IM subspecialty diagnoses as pediatricians, and identified the physicians of adults with these diagnoses as general/family practitioners or internists in either general internal medicine or an IM subspecialty. b The imputation method identified no patients with this characteristic for this specialty.

	Under a status quo scenario where per capita patterns of health care use within a defined demographic group are assumed to remain constant over time, future demand for health care services can be extrapolated by estimating the size of the population in each demographic group and applying the corresponding per capita utilization rates. Analyses to update the NDM found that under such a scenario the growth and aging of the population between 2000 and 2020 would contribute to a 30 percent increase in inpatient days at general, short-term hospitals; a 20 percent increase in non-emergency outpatient visits to hospitals; a 33 percent increase in inpatient days at non-general and long-term hospitals; a 17 percent increase in emergency department visits; a 36 percent increase in home health visits; and a 40 percent increase in nursing home residents. Estimates from the PARM suggest that visits to physician offices would increase by 23 percent under this status quo scenario. Although recent history is often the best predictor of future health care utilization rates, many analysts argue that future rates might differ from current patterns because: The needs of the population are changing (even after controlling for demographics); The health care operating environment is constantly changing; Economic considerations may make current utilization trends unsustainable as the size of the elderly population increases; New diseases could emerge; and Technological advances will change how and where services are provided. A detailed analysis of the impact on the future health workforce of changes to the health care operating environment and technological advances is beyond the scope of this effort; however, Section 5 contains forecasts from the PARM and NDM for scenarios that rely on different assumptions regarding the future health care operating environment and other determinants of the demand for health care providers. A report entitled: The Impact of the Restructuring of the U.S. Health Care System on the Physician Workforce and on Vulnerable Populations (The Lewin Group, 1998) examines several emerging trends in the health care system and discusses their implications for the future physician workforce. The impact of advances in science and medicine on demand for health care services and the productivity of health care providers will differ by medical specialty and delivery setting. Advances could increase workforce demand in some settings or specialties while decreasing demand in other settings or specialties. For example, technological advances are making outpatient surgery a viable alternative to inpatient surgery, and this is contributing to the decrease in inpatient days and the increase in outpatient visits. Yashar (2000) reports that improvements in surgical instruments have transformed how ocular surgery is performed and that ambulatory surgery is becoming the norm for most ocular surgery.

	Similarly, Balaban (1998) states that technological improvements and efforts to contain costs have contributed to the trend where bone marrow transplants are performed on an outpatient basis with following-up ambulatory visits. Gelijns and Fendrick (1993) provide other examples such as cholecystectomy and cardiac catheterization where minimally invasive surgical procedures have shifted many of these procedures from an inpatient to an outpatient setting. This trend is occurring in many medical specialties and is likely to continue over the next few decades. Hospitalization will still occur when treating the more severe cases; consequently, while total inpatient days are expected to decline at acute care hospitals, average patient acuity is likely to rise and this could affect staffing patterns. In addition, the development of new medications could also reduce future demand for some health care services, and thus demand for some health professionals. Advances in science and medicine are contributing to higher life expectancy. Over the past 100 years, life expectancy has doubled. Increased longevity will contribute to greater demand for health care over the long run. An important question for projecting future demand for health professionals as the population ages is whether current utilization rates for the elderly accurately represent future utilization rates for that group. Much of this debate centers on the oldest elderly, who have the highest per capita utilization of health care services. In addition to advances in science and medicine and improvements in public health, there are important differences between today's elderly and tomorrow's elderly that could lead to lower per capita utilization in the future. These differences include changes in lifestyle of the rising elderly cohort, such as improved diet and exercising, higher educational attainment, and greater economic resources. The consensus is that higher education and greater economic resources, which are highly correlated, will contribute to improvements in the health status of the rising elderly cohort because both education and economic resources contribute to a healthier lifestyle. Greater economic resources allow individuals to purchase the inputs to better health via more nutritious food, increased or better preventive care, improved information, and more effective pharmaceuticals. Freedman and Martin (1998) find that better educated elderly are more likely to comply with physicians' instructions, which leads to fewer complications. Manton, Corder, and Stallard (1997) find that people with higher levels of education are less likely to be disabled when controlling for age and other characteristics.

	The extant literature finds that disability rates among the elderly have been declining slightly, resulting in a decline in use of some health care services. Bishop (1999) reports that per capita use of nursing home services has declined over the past decade. Possible explanations cited include lower disability rates among the elderly, the rise in alternative health care services such as home- or community-based care, economic considerations, changes in the health care operating system, changes in government programs such as Medicare and Medicaid, and other factors cited above. Manton et al. (1997) find that disability rates among older Americans are declining slightly. Using data from the National Long-term Care Surveys, these authors find that in 1994 an estimated 21.3 percent of the age 65 and older population were chronically disabled. If disability rates had remained at their 1982 levels, an estimated 24.9 percent of older Americans would have been chronically disabled in 1994, an imputed difference of 3.6 percentage points. Freedman and Martin (2000) used data from the Supplements on Aging to the 1984 and 1994 National Health Interview Surveys to examine trends in chronic conditions and functional limitations of Americans 70 years and older. They report that the percentage of older Americans with functional limitations relating to seeing, lifting, carrying, climbing, and walking declined between 1984 and 1994. Bonifazi (1998) analyzed the number and needs of nursing home residents in 1995 compared to 1977. He finds that a smaller percentage of older Americans are entering nursing homes-41 per thousand in 1995 compared to 47 per thousand in 1977-despite the aging of the elderly population. Part of this decline is attributed to the increase in alternative care settings such as outpatient care and home health care. Declining disability rates among the elderly could help reduce the projected high growth in demand for nursing home care. In addition, the growth in community-based care could further reduce per capita demand for institutionalized care. As elderly with less severe health problems opt out of nursing homes for home- and community-based care, the health care needs of the average nursing home resident rises. Hence, future demand for nurses and other health workers in nursing homes could rise proportionately faster than the growth in nursing home residents as the population ages. In community-based settings, the impact of declining disability rates is unclear. On the one hand, declining disability rates might decrease demand for services. On the other hand, declining disability rates could shift care from an institutional setting to a community- or home-based setting. Alecxih (2001) finds that the increase in the size of the elderly population will likely overwhelm other factors that might influence the future demand for medical care from the elderly. Alecxih examined the potential impact of socioeconomic trends on demand for long-term care, including declining disability rates, increased availability of informal support networks, and a more highly educated elderly cohort. She estimates that demand for long-term care will more than double by 2050 because of the increasing size of the elderly population. Stuki and Mulvey (2000) estimate that by 2030, when the last of the baby boomers reach age 65, an estimated 6 million elderly could be at risk of institutionalization because of severe impairments.

	Although the literature suggests numerous factors that could reduce per capita demand for health care services from tomorrow's elderly compared to today's elderly, Glied and Stabile (1999) provide an example of one factor that could cause health care utilization rates for the elderly to rise in coming years. These authors predict that private insurance coverage among the near-elderly (i.e., persons ages 61-64) will drop by 4.5 percent by 2005 because of trends relating to the labor market behavior of the elderly and the reduced propensity of employers to offer medical insurance. Although the proportion of the population age 61 to 64 employed full time increased between 1989 and 1997, the authors report that older workers have been affected by the nationwide decline in private medical insurance coverage. The leading edge of the baby boom generation is just now entering the phase where they are not yet eligible for Medicare and are, for the most part, relying on their current or past employer (if retired) to obtain medical insurance. Declining rates of medical coverage among the near-elderly could result in a decline in preventive care with long-term implications for this group as they age.
	2.2.2 Increasing Demand for Health Workers
	Who will provide for the health care needs of the future elderly and where will they receive care? Currently, the elderly are cared for by services paid for by Medicare, Medicaid, private insurers, and out-of-pocket. In addition, many elderly rely on an informal network of unpaid workers-usually family members. Several demographic trends could change the mix of people and institutions providing care to the elderly. As discussed above, declining disability rates among the elderly, controlling for age, might allow more elderly to remain in their homes or in other community-based settings. This would place fewer demands on providers of institutional care, but would increase demand for home-based services provided by home health aides, nurses, physical therapists, and other paid professionals. This could also increase demand for unpaid providers even while several trends suggest that in the future the elderly will have a smaller network to rely on for informal, long-term care. Consider the following factors that could reduce the future supply of unpaid health care providers. First, increased longevity means that the adult children of some elderly will themselves be elderly. In future years, it might be common for a 70-year old to care for his or her 90-year-old parent. The physical demands of caring for a disabled parent might be too great for many elderly children, which could increase demand for home- and community-based care as well as institutionalized care. Second, baby boomers had relatively small families compared to earlier generations, so they will have fewer children to provide unpaid care than today's elderly. Third, Stuki and Mulvey (2000) note that baby boomers had higher divorce rates than today's elderly, and research by Schone and Pezzin (1999) finds that divorced parents are less likely than widowed parents to receive long-term care from their adult children. Fourth, women traditionally have provided the bulk of unpaid care for elderly parents and the proportion of women in the workforce has increased during recent decades. Providing long-term care to an elderly parent or family member might require many of these women (or men) to leave the workforce or to reduce the number of hours worked. An estimated 40 percent of people who provide care to a severely-impaired, older parent or family member are employed, and a significant number of these caregivers are forced to adjust their work schedule or take a leave of absence (NAC and AARP, 1997). A higher proportion of women in the workforce makes it more expensive for family members to care for their disabled parents or relatives, but also makes it financially easier to purchase services from home health agencies and institutional care providers.

	As the aging population demands more health care services, the demand for health workers will increase. Demand will grow faster for those specialties that disproportionately serve the elderly population. For example, Angus et al. (2000) discuss the implications of the growing elderly population on projected demand for physicians in adult critical care and pulmonary medicine. The authors report that two-thirds of all inpatient pulmonary days are incurred by patients age 65 and older. The projected growth in demand for services in these areas leads the authors to predict a growing shortage of physicians in adult critical care and pulmonary medicine during the next two decades. Using the PARM, one can estimate the proportion of time physicians spend with patients in different age groups. In this model, as discussed previously, the average length of time that physicians spend per visit with patients in physicians' offices and hospital outpatient settings varies by patient demographic characteristics and insurance status. In the other settings modeled in the PARM, the assumption is made that physician time per encounter is independent of patient age, sex, race/ethnicity, and insurance status. Currently, physicians spend an estimated 16 percent of patient-care hours providing services to children under age 17, 15 percent with the age 18-34 population, 26 percent with the age 34-54 population, 11 percent with the age 55-64 population, 14 percent with the age 65-74 population, and 18 percent with the age 75 and older population (Exhibits 2.11 and 2.12). These estimates combine differences in health care needs and size of the population in each age group, as well as differences in physician time per visit in settings where this information is available. As expected, the proportion of time physicians spend with elderly patients will increase as the population ages and the elderly comprise a larger proportion of the population. Consider a scenario where physician productivity, staffing levels, and health care use patterns within a demographic group remain constant over time at their 1999 levels. In 2020, physicians would be spending an estimated 39 percent of total patient-care hours providing services to the age 65 and older population compared to an estimated 32 percent in 2000. Today, the 35-54 age group, which closely corresponds with the baby boom generation, consumes an estimated 26 percent of total patient-care hours. In 20 years, baby boomers will be in the 55-74 age group and will consume approximately 34 percent of total patient-care hours. The impact of the increasing age of the population on the percentage of total patient care hours spent with each age group is shown below for physicians in general primary care (Exhibit 2.13), other medical specialties (Exhibit 2.14), surgery (Exhibit 2.15) and other patient care (Exhibit 2.16).

	Exhibit 2.11. Estimated Percentage of Physician's Time Spent Providing Care to Patients, by Age of Patient
	Specialty Age Category, 2000 Age Category, 2020 1-17 18-34 35-54 55-64 65-74 75 + 1-17 18-34 35-54 55-64 65-74 75 + Total Patient Care Physicians (MDs and DOs) 16 15 26 11 14 18 14 12 20 15 19 20 General Primary Care 29 11 22 10 12 16 25 9 17 14 16 19 General & Family Practice 13 16 29 12 13 17 10 14 22 17 18 20 General Internal Medicine 3 11 28 14 18 25 2 9 20 19 23 27 Pediatrics 100 0 0 0 0 0 100 0 0 0 0 0 Medical Specialties 6 10 26 15 20 23 5 8 19 19 25 24 IM Subspecialties 2 9 28 16 21 23 2 7 20 20 26 24 Cardiovascular Diseases 2 3 18 17 27 33 1 2 12 20 33 32 Other Medical Specialties 17 16 29 12 12 15 14 13 23 16 16 18 Surgery 7 23 27 11 15 17 6 20 20 16 20 19 General Surgery 7 12 27 15 19 20 6 10 20 19 24 22 Obstetrics/Gynecology 4 56 31 5 3 2 4 55 27 7 5 2 Otolaryngology 22 15 27 10 11 14 19 14 21 15 16 16 Orthopedic Surgery 9 14 25 12 16 24 7 11 18 16 21 27 Urology 6 9 21 15 23 26 4 7 14 18 29 27 Ophthalmology 9 7 20 14 22 28 7 6 14 17 28 28 Other Surgical Specialties 4 9 29 16 21 21 3 7 20 21 27 22 Other Patient Care 11 16 31 11 13 18 9 13 24 15 18 21 Anesthesiology 10 18 43 10 8 11 9 16 35 15 11 14 Emergency Medicine 11 15 20 12 19 24 9 12 15 15 24 25 Radiology 24 27 26 7 7 8 22 25 22 11 10 10 Pathology 13 12 21 11 17 26 10 9 16 15 21 28 Psychiatry 1 10 42 12 16 18 1 8 33 16 21 20 Other Specialties 6 15 33 12 14 20 5 12 25 16 19 22 Non-Physician Specialties Physical Therapy 20 17 33 12 10 9 17 15 26 17 14 10 Optometry 16 19 31 12 11 10 14 17 24 17 16 11 Podiatry 9 13 29 14 16 20 7 11 22 19 21 21 Total U.S. Population (Thousands) 26 24 29 9 7 6 24 22 25 13 10 7
	Source: These forecasts from the Physician Aggregate Requirements Model assume no change over time in per capita utilization, physician productivity or mix, or the health care operating environment. Note: percentages might not sum to 100 percent due to rounding.
	Exhibit 2.12: Distribution of Total Patient Care Hours, by Patient Age: Total Active Physicians in Patient Care
	Exhibit 2.12: Distribution of Total Patient Care Hours, by Patient Age: (Text Only) Total Active Physicians in Patient Care
	0-17 18-34 35-54 55-64 65-74 75 + 2000 16% 15% 26% 11% 14% 18% 2020 14% 12% 20% 15% 19% 20%
	Exhibit 2.13: Distribution of Total Patient Care Hours, by Patient Age: General Primary Care Physicians
	Exhibit 2.13: Distribution of Total Patient Care Hours, by Patient Age: General Primary Care Physicians (Text Only)
	0-17 18-34 35-54 55-64 65-74 75 + 2000 29% 11% 22% 10% 12% 16% 2020 25% 9% 17% 14% 16% 19%
	Exhibit 2.14: Distribution of Total Patient Care Hours, by Patient Age: Primary Care Subspecialty Physicians
	Exhibit 2.14: Distribution of Total Patient Care Hours, by Patient Age: Primary Care Subspecialty Physicians(Text Only)
	0-17 18-34 35-54 55-64 65-74 75 + 2000 6% 10% 26% 15% 20% 23% 2020 5% 8% 19% 19% 25% 24%
	Exhibit 2.15: Distribution of Total Patient Care Hours, by Patient Age: Physicians in Surgical Specialties
	Exhibit 2.15: Distribution of Total Patient Care Hours, by Patient Age: Physicians in Surgical Specialties (Text Only)
	0-17 18-34 35-54 55-64 65-74 75 + 2000 7% 23% 27% 11% 15% 17% 2020 6% 20% 20% 16% 20% 19%
	Exhibit 2.16: Distribution of Total Patient Care Hours, by Patient Age: Physicians in Other Patient Care Specialties
	Exhibit 2.16: Distribution of Total Patient Care Hours, by Patient Age: Physicians in Other Patient Care Specialties (Text Only)
	0-17 18-34 35-54 55-64 65-74 75 + 2000 11% 16% 31% 11% 13% 18% 2020 9% 13% 24% 15% 18% 21%
	2.3 Implications of an Aging Population for the Supply of Health Workers
	Demographic trends in the health workforce will mirror many of the trends in the overall population. In many health care occupations, there are a significant number of baby boomers that will retire just as demand for their services is increasing. This is especially true in nursing. An emerging nursing shortage is likely to be exacerbated starting in approximately 2010 as a large portion of the nurse workforce nears retirement. In occupations where some analysts argue there is a current surplus-e.g., specialist physicians-the growth in demand for services and retirement from the physician workforce could eliminate any surplus and could even result in shortages. A large majority of the relevant workforce supply literature focuses on physicians and registered nurses, with much less published on other health workers.
	2.3.1 Physician Supply
	Forecasting the future supply of physician services involves attempting to predict the future rate of entrance to and exit from the profession, and predicting the productivity of these physicians while they are in the workforce. The age distribution of both the U.S. population and the current physician workforce is an important determinant of the size and characteristics of the future workforce. The age distribution of the U.S. population affects the rate of new entrants to the profession, while the age distribution of the physician workforce affects rates of exit and average level of physician productivity. Productivity is defined here as the average number of patient hours per physician per year. Physicians, like many professionals who invest heavily in their training, remain active in their professions throughout a working career of 30 or more years. The literature suggests that the rate at which physicians exit the workforce or reduce their workload is highly related to age-especially as physicians approach retirement age. American Medical Association (AMA) publications show the number of active physicians in different age groups. Of those physicians under 65 years of age in the AMA MasterFile in 1999, 18.9 percent were under age 35, 32.4 percent were age 35-44, 31 percent were age 45-54, and 17.8 percent were age 55-64 (Exhibit 2.17). The age distribution varies substantially by reported primary medical specialty, possibly reflecting when a specialty was officially founded (Exhibit 2.18). For example, 47.4 percent of general practitioners and 40.1 percent of radiologists were age 55-64, while only 10 percent of emergency physicians and 10.5 percent of family practitioners were in this age group. In thoracic surgery, approximately half the physicians are under age 35 and the other half are almost entirely age 35-44. There are very few physicians over age 44 who report thoracic surgery as their primary specialty. Some specialties, such as general surgery, have a relatively flat age distribution, with approximately 1/4th of physicians in each of the four age groups. Specialties with a high percentage of physicians nearing retirement are especially vulnerable to a rapid decrease in number of active physicians. Not only is an adequate supply of new physicians important to consumers, but an adequate supply is important to retiring physicians who desire to see established practices continue to flourish.
	Exhibit 2.17. Age Distribution of the Current Physician Workforce
	Exhibit 2.17. Age Distribution of the Current Physician Workforce (Text Only)
	0-35 35-44 45-54 55-64 19% 32% 31% 18%
	Source: American Medical Association, Physician Characteristics and Distribution in the U.S., 2001-2002 Edition.
	Exhibit 2.18. Percent Distribution of the Physician Workforce Under Age 65, by Age Group, in 1999
	Specialty Under 35 Years 35-44 Years 45-54 Years 55-64 Years Total 18.9 32.4 31.0 17.8 Aerospace Medicine 2.9 28.5 41.5 27.1 Allergy & Immunology 5.8 31.4 38.0 24.8 Anesthesiology 13.2 42.4 29.2 15.2 Cardiovascular Disease 9.4 35.5 36.1 18.9 Child Psychiatry 8.4 34.5 36.3 20.8 Colon/Rectal Surgery 7.8 36.5 35.8 19.9 Dermatology 16.7 31.0 31.6 20.7 Diagnostic Radiology 18.7 34.2 30.9 16.2 Emergency Medicine 23.0 31.9 35.0 10.0 Family Practice 22.9 34.7 31.8 10.5 Forensic Pathology 6.6 32.3 35.7 25.4 Gastroenterology 9.6 37.3 35.8 17.3 General Practice 1.1 13.9 37.6 47.4 General Preventive Med. 7.2 31.0 37.4 24.3 General Surgery 24.1 26.9 26.9 22.1 Internal Medicine 24.8 33.0 29.5 12.7 Medical Genetics 13.2 31.9 34.0 20.8 Neurology 12.6 32.7 36.5 18.3 Neurological Surgery 18.6 28.7 27.6 25.1 Nuclear medicine 9.0 24.7 37.7 28.6 Obstetrics/Gynecology 19.2 30.1 30.6 20.1 Occupational Med. 0.6 27.7 45.2 26.5 Ophthalmology 12.6 31.5 31.2 24.7 Orthopedic Surgery 17.1 29.8 29.7 23.4 Otolaryngology 18.0 29.8 27.4 24.8 Pathology-Anat/Clin 12.2 31.7 32.6 23.6 Pediatrics 27.0 32.3 27.0 13.7 Pediatric Cardiology 13.9 42.2 26.6 17.3 Physical Med/Rehab 18.2 42.7 26.3 12.8 Plastic Surgery 7.5 32.2 35.1 25.1 Psychiatry 11.2 28.1 34.7 26.0 Pulmonary Diseases 11.3 36.5 37.3 14.9 Radiology 7.8 27.1 25.0 40.1 Radiation Oncology 13.6 37.8 29.5 19.0 Thoracic Surgery 50.4 49.2 0.4 0.0 Urological Surgery 13.5 27.6 29.9 29.0 Other 1.8 20.1 40.1 38.1
	Source: American Medical Association, Physician Characteristics and Distribution in the U.S., 2001-2002 Edition
	As health professionals age, they typically reduce their hours worked in patient care-especially professionals approaching retirement age who might view a reduced workload as an alternative to retirement. Although we identified no recent studies showing working patterns of physicians over their career, a survey of optometrists by Abt Associates (White, Doksum and White, 2000) finds that hours spent in patient care decline with age (Exhibit 2.19). The trend is especially evident among male optometrists. From age 30 to retirement, average hours spent in patient care drops slowly but steadily. Average hours worked by female optometrists declines slightly when these women are in their 30s and 40s, possibly resulting from a reduced workload to care for children, but then increases in their 50s until retirement. The spike in hours by female optometrists in the 65-69 age group could be an anomaly due to small sample size.
	Exhibit 2.19. Average Number of Hours Optometrists Spend in Patient Care per Work Week
	Hours Spent in Patient Care Age Group Men Women 25 to 29 41.6 40.4 30 to 34 43.0 37.5 35 to 39 42.3 35.6 40 to 44 41.7 34.2 45 to 49 41.2 35.4 50 to 54 39.8 37.1 55 to 59 38.6 37.0 60 to 64 37.2 35.2 65 to 69 33.3 42.3 70+ 28.5 27.1
	Source: Project Hope Census of Optometrists (White, Doksum and White, 2000), Table 2.
	2.3.2 Nurse Supply
	The aging of the nurse workforce and the inability to attract new entrants are often cited as major contributors to an impending nurse shortage. [5] The extant literature on this topic is vast, but two recent publications include a study by the General Accounting Office (2001) and Nevidjon and Erickson (2001). Factors contributing to the aging of the nurse population include the large number of baby boomers who entered the profession in the 1970s and 1980s, declining enrollment in nursing programs, retention difficulties, and a higher average age of new graduates from nursing programs. Findings from the 2000 Sample Survey of Registered Nurses (HRSA, 2001) indicate that between 1980 and 2000 the percentage of RNs under the age of 40 fell from approximately 53 percent to 32 percent. Buerhaus, Staiger and Auerbach (2000) discuss this phenomenon and the implications of an aging RN workforce. The authors report that between 1983 and 1998 the average age of RNs in hospitals increased by 5.3 years. During the same period, the average age of the entire RN workforce increased 4.5 years, from 37.4 to 41.9. The General Accounting Office (GAO, 2001) estimates that by 2010, approximately 40 percent of the RN workforce will be age 50 or older.
	The primary cause of an aging RN workforce is the failure to attract young workers (especially women) into the profession. The changing age distribution of the population will make it more difficult to attract young workers into nursing in future years. The American Association of Colleges of Nursing reports that enrollments in entry-level baccalaureate programs in nursing have declined every year between 1995 and 2000. Enrollees to these programs have declined by 21 percent between 1995 and 2000, while graduates have declined by 16.5 percent. The GAO estimates that the ratio of working-age women (age 18 to 64) to the age 85 and older population will decline over time from approximately 40:1 in 2000, to 22:1 in 2030, and to 15:1 in 2040. This finding has important implications for the future supply of all health professions. Buerhaus, Staiger and Auerbach analyzed the relationship between age and RN workforce participation for a cohort (defined by birth year) of the population. RNs typically enter the profession in their early 20s to early 30s, and the number of full-time equivalent (FTE) RNs from a population cohort increases through age 45 as many RNs finish their schooling and pass out of their child rearing years. Between ages 45 and 55, the number of FTEs from a cohort remains fairly stable, but then begins to decline as RNs retire or reduce hours worked. Although the demographics of the current nurse workforce will have a great impact on the nurse workforce of the future, the large proportion of nurses who will be retiring during the next 10 years will not necessarily result in a shortage. Economic theory suggests, and history has shown, that wages will adjust, making shortages and surpluses a short-term phenomenon. However, it does suggest that the real wages of nurses will increase. This in turn will attract new entrants, gradually reducing wages to "normal" levels. There is less literature on the demographics of licensed practical nurses and nurse aides. LPNs and nurse aides tend to be younger than RNs. Indeed many LPNs and nurse aides see becoming RNs as a means to better oneself professionally. The duties performed by LPNs and nurse aides are often physically demanding which limits the ability of some older people to serve in this capacity. Because LPNs and nurse aides require less time to train than RNs, the supply of these nurses can react more quickly to market conditions. As an aging population demands more services from an increasingly older nurse workforce, some employers of nurses might look outside the U.S. to countries with younger populations. Many of these countries that could potentially export nurses might themselves have nurse shortages, in which case an inadequate supply of nurses in the U.S. could reduce the availability of care in other countries. Cheryl Peterson, director of international nursing at the American Nurses Association, states: "I'm always telling people in developing countries, 'You don't want the U.S. shortage to worsen because we'll grab up all of the world's poor nurses.'" [6] As reported in the Wall Street Journal article: Shortage of Nurses Hits Hardest Where They Are Needed the Most: Nurse Shortage Shows How Labor Markets Go Global (Zachary, 2001, p. A12).

	2.4 Implications of an Aging Population for the Economics of the Health Care System
	Health care spending constitutes almost one-eighth of our Gross Domestic Product (Heffler, 2001). Because such a large portion of the Nation's resources is spent on health care, the economics of the health care system are closely intertwined with the national economy. Changing demographics will have a significant impact on both the U.S. economy and the economics of the health care system. The Congressional Budget Office (1997) estimates that total national spending on health care could double between 1996 and 2008 to nearly $2 trillion. Ginzberg (1999) projects that annual expenditures for health care could top $4 trillion by 2025, and this, says Ginzberg, "could turn out to be a serious underestimate given the steep increase in the number of elderly, who make much greater use of health care services than the below-65 population (p. 58)." Stucki and Mulvey (2000) report that by 2030, when the last of the baby boomers reaches age 65, the cost to provide personal care, adult day care, and assisted living to the elderly could quadruple to an estimated $193 billion. Nursing home expenditures paid by Medicaid could rise 360 percent to $134 billion (in 1996 dollars) between 2000 and 2030 (Mulvey and Stucki, 1998). If retirement patterns remain unchanged, the ratio of working to retired Americans will continue to decline as the population ages. Pizer, Frakt and Kidder (2000) project that by 2005 the ratio of workers to retirees will be 5:1, and this ratio could fall to 2.75:1 by 2050. This means that a smaller proportion of the population will be supporting the needs of the elderly. The Medicaid and Medicare programs will compete with other programs, such as Social Security, that serve the elderly. As the size of the elderly population grows, resulting in an increase in the number of Medicare and Medicaid eligibles, the resulting increase in government outlays for health care services could compel the government to reduce expenditures by reducing benefit levels, restricting eligibility, increasing out-of-pocket expenditures by increasing premiums or co-pays, and reducing reimbursements to health care providers. On the other hand, the elderly will constitute a growing voting bloc that could attempt to retain current benefits or even expand benefits. Tarlov (1995) states that the consensus outlook of future demand for health care services is that "service quantity and price will be set at economically absorbable levels determined by employer-employee willingness to pay and by politically acceptable government budgets for health care (p. 1560)." Ginzberg anticipates that cost pressures will result in radical changes in the health care system during the early part of the 21st century. Ginzberg anticipates that Medicare will provide beneficiaries access to "essential" health care services, but not to high-cost hospitals and expensive procedures.

	Actions to reduce spending could reduce demand for health workers. The impact would vary substantially by medical specialty and delivery setting, with providers of expensive services likely to see the greatest impact on demand for their services. In addition, attempts to reduce health care spending through lower reimbursement rates to health care providers could, in the long run, reduce the supply of health workers. Caro and Kaffenberger (2001) find that reductions in Medicare payments for nursing home care and home health services resulting from the Balanced Budget Act of 1997 pushed many long-term care providers out of business, thus reducing the demand for nurses and other health workers in those settings.
	CHANGING RACIAL AND ETHNIC COMPOSITION OF THE POPULATION
	Advocates for increased minority representation in the health workforce argue that increasing the number of minority physicians will improve access to care for minorities and vulnerable, underserved populations. Major Findings: Minorities have different patterns of health care use compared to non-minorities. Disparities in access to care account for part of the difference in utilization. Demand for health care services by minorities is increasing as minorities grow as a percentage of the population. Between 2000 and 2020, the percentage of total patient care hours physicians spend with minority patients will rise from approximately 31percent to 40 percent. Minorities are underrepresented in the physician and nurse workforce relative to their proportion of the total population. As minorities constitute a larger portion of the population entering the workforce, their representation in the physician and nurse professions will increase. The U.S. will increasingly rely on minority caregivers. Minority physicians have a greater propensity than do non-minority physicians to practice in urban communities designated as physician shortage areas. An increase in minority representation in the physician workforce could improve access to care for the population in some underserved areas. These advocates argue that increased representation of minorities in the health workforce not only will increase equity, but will also improve the efficiency of the health care delivery system. This section explores the changing racial and ethnic composition of the population and its implications for the future demand for and supply of health professionals. The four main findings are the following. First, Hispanics and non-whites have different patterns of health care use compared to non-Hispanic whites. Some of the disparities in use can be attributed to differences in access to care. The literature suggests that cultural differences regarding appropriate use of health care services also help explain differences in health care use. Second, as minorities increase as a percentage of the U.S. population, the percentage of total health care services provided to minority patients will also increase. In 2000, physicians spent an estimated 31 percent of patient-care hours providing services to minorities. By 2020, physicians will spend an estimated 40 percent of patient-care hours with minority patients.

	Third, minorities are underrepresented in the physician and nurse workforces relative to their proportion of the total population, and are overrepresented in lower-paying health professions such as nurse aides and home health aides. As minorities constitute a growing percentage of the working-age population, their representation in the professional health workforce will naturally rise. The U.S. will increasingly rely on minority caregivers. Fourth, the literature suggests that minority physicians have a greater propensity than do non-Hispanic white physicians to practice in urban communities designated as physician shortage areas. An increase in minority representation in the physician workforce could improve access to care for the population in some underserved areas.
	3.1 Population Forecasts
	The latest census figures highlight the fact that the United States is becoming increasingly racially and ethnically diverse. Furthermore, higher birth rates among racial and ethnic minority groups, relative to non-Hispanic whites, and immigration suggest that this trend will continue. Exhibit 3.1 contains population forecasts used in the PARM that show the current and projected distribution of the population across the three race/ethnic groups modeled in the PARM. Whereas non-Hispanic whites constituted approximately 69 percent of the population in 2000, they will constitute an estimated 61 percent of the population in 2020. Between 2000 and 2020, African Americans (both Hispanic and non-Hispanic) will increase from approximately 12.3 percent to 13.1 percent of the population; all other minorities (including Hispanic whites) will increase from approximately 19 percent to 26 percent of the population. Growth in the Hispanic population is the major contributor to growth in the minority population.
	Exhibit 3.1. Population Distribution by Race
	Year Non-Hispanic White African American All Other 2000 69.1% 12.3% 18.6% 2005 67.1% 12.5% 20.4% 2010 64.8% 12.7% 22.5% 2015 62.8% 12.9% 24.3% 2020 60.8% 13.1% 26.1%
	Source: Modified version of Census Bureau middle series projections.
	Racial and ethnic minority populations are unevenly distributed geographically. The proportion of a State's population that is minority varies substantially by State, and minorities are disproportionately located in inner cities.

	3.2 Implications of the Changing Racial and Ethnic Composition of the Population for the Demand for Health Workers
	The extant literature explores the degree to which and reasons why race and ethnicity may affect health care use. Differences between racial and ethnic groups in use of a wide range of health care services have been documented in the literature. Much of these utilization differences are attributed to differences in access to care and cultural differences regarding the use of health care services. A better understanding of differences in health care utilization by race and ethnicity, the causal factors of these differences, and whether these differences will persist in the future allows for better predictions of future demand for health workers. Below is a sample of the literature that describes differences in health care utilization by race or ethnicity. Mueller, Patil and Boilesen (1998) analyzed data from the 1992 National Health Interview Survey (NHIS) and found racial disparities in use of physician services even after controlling for factors such as insurance status, geographic location and other patient characteristics. The disparity in use of physician services by race was not statistically different from zero for those patients living in urban areas, but the disparity was statistically different from zero for patients living in rural areas. Insurance status and location (urban versus rural) are greater determinants of use of physician services than is patient race. Hargraves, Cunningham and Hughes (2001) found small differences in access to care and health care use of non-Hispanic whites and minorities enrolled in managed care plans. Whereas approximately 78 percent of non-Hispanic whites have a regular provider, only 74 percent of Hispanics and African Americans have a regular provider. Whites have slightly higher use of specialists. In their last physician visit, 28 percent of non-Hispanic whites saw a specialist compared to 26 percent for African Americans and 22 percent for Hispanics. Burns et al. (1996) use Medicare claims from ten States to examine differences in mammography use between elderly African American and white women. They find that African American women had lower use rates than white women across all levels of primary care. These authors cite additional research that finds that physicians are more likely to encourage elderly white women to obtain mammograms than elderly African American women, highlighting concerns about provider attitudes. Peterson et al. (1994) analyzed the use of cardiac procedures of men treated at Veterans Affairs Medical Centers. These authors find that African Americans are less likely than their white counterparts to undergo selected cardiac procedures. The authors suggest several reasons for the differences in treatment, including: (1) differences in severity, (2) consumer preferences, and (3) differences in how providers may weigh the risk and benefit of invasive procedures differently for African Americans than for whites. Todd et al. (1993) studied analgesic use in emergency departments and find that ethnicity was a strong predictor of the lack of use of analgesics. Mitchell et al. (2000) analyzed Medicare inpatient data to compare differences between African Americans and whites in the use of diagnostic and therapeutic services for cerebrovascular disease. These authors control for differences in factors such as health care needs and ability to pay. Still, they find that "black patients were significantly less likely to receive non-invasive cerebrovascular testing, cerebral angiography, or carotid endarterectomy compared to white patients (p. 1413)."

	Not all studies find differences by race or ethnicity in use of health care services. For example, Horner et al. (1997) found no differences by race and ethnicity in the use of inpatient rehabilitation services for elderly stroke victims after adjusting for differences in patient risk. Access to affordable medical insurance is often cited as a major determinant of access to care. People in racial and ethnic minority groups in 1999 were more than twice as likely as nonminorities to be uninsured. The Census Bureau estimates that, in 1999, 89 percent of non-Hispanic whites had some form of medical insurance while only 79 percent of African Americans and 67 percent of Hispanics were insured. [7] US Census, http://www.census.gov/hhes/hlthins/hlthin99/hi99tc.html These statistics are important because the literature has established a link between access to care and health status (e.g., Drake and Lowenstein, 1998). Specifically, people without medical insurance tend to receive less preventative care and have higher rates of hospitalization for potentially avoidable problems. Drake and Lowenstein note that in California during the year of their study (1993), approximately 14 percent of African Americans and 37 percent of Latinos were uninsured, compared to 12.5 percent of whites. An analysis of the 1999 NHIS found that 9 percent of non-Hispanic whites, 16.4 percent of African Americans, and 26.3 percent of other minorities (including Hispanic whites) were without health insurance on the date surveyed in 1999. The PARM divides the population into three insurance categories: insured in a fee-for-service arrangement, insured in an HMO, and uninsured. Exhibit 3.2 shows that the proportion of each racial/ethnic group in an HMO is relatively similar, controlling for age and sex, but the percentage insured in a fee-for-service arrangement and uninsured vary substantially by race/ethnicity. Language and cultural differences also are cited as factors affecting health care utilization. With the growing population of Hispanics in the U.S. and immigration from non-English speaking countries, language is playing an increasingly important role in the provision of health care services. Consider the following findings in recent studies. Kravitz et al. (2000) found that Spanish-speaking patients who visited the General Medicine and Family Practice Clinics at the UC Davis Medical Center were less likely to follow up with recommended laboratory studies compared to English-speaking patients. In addition, patients needing a translator required more physician time per visit. The authors applied regression models to estimate the impact of language on physician time per visit. They found that Spanish- and Russian-speaking patients averaged 9.1 and 5.6 additional minutes of physician time, respectively, compared to English-speaking patients after controlling for other determinants of physician time per visit.

	Exhibit 3.2. Percent Distribution of the Population by Demographic Group Across Three Insurance Categories
	Non-Hispanic White African American All Other Age Insurance Male Female Male Female Male Female 0-17 FFS 59 59 59 61 48 48 HMO 34 34 30 27 29 29 Uninsured 7 7 11 12 23 23 18-34 FFS 48 52 39 46 30 35 HMO 32 33 29 33 28 31 Uninsured 21 15 32 21 42 34 35-54 FFS 55 56 44 45 35 39 HMO 35 35 37 36 36 36 Uninsured 11 9 19 18 29 25 55-64 FFS 61 63 57 59 43 48 HMO 32 30 28 25 37 31 Uninsured 7 7 15 16 20 21 65-74 FFS 85 87 82 82 85 85 HMO 15 13 18 18 15 15 Uninsured 0 0 0 0 0 0 75+ FFS 89 89 87 93 88 93 HMO 11 11 13 7 12 7 Uninsured 0 0 0 0 0 0 All Ages FFS 60 63 52 56 41 46 HMO 30 29 30 29 30 30 Uninsured 10 8 18 15 28 24 FFS 61 54 44 HMO 30 30 30 Uninsured 9 16 26
	Source: Analysis of the 1999 NHIS.
	Derose and Baker (2000) analyzed survey data for 465 Spanish-speaking Latinos and 259 English speakers of various ethnicity who presented to a public hospital emergency department in Los Angeles. The survey asked participants to assess their English-speaking ability; indicate the number of visits to a physician during the prior three months; and provide information on the participants' health status, socioeconomic status, and demographic characteristics. The authors found that of participants who had at least one visit to a doctor during the previous three months, those with limited English proficiency had 22 percent fewer visits, on average, compared to participants with good-to-excellent English proficiency. The study controlled for patient characteristics that could be correlated with the use of physician services such as health conditions and insurance status. In practice, therefore, language and communication may be significant barriers to access to care. In addition to differences across racial groups and English/non-English speakers in access to and use of health care services, there are significant differences in measures of health status that affect the type of care demanded. Keppel, Pearcy, and Wagener (2002) find that compared to non-Hispanic whites, many minority populations have higher infant mortality rates, higher rates of infants with low birth weight, higher age-adjusted rates of heart disease death, higher rates of tuberculosis, and disparities in many other measures of health care.

	Freiman (1998) argues that the relationship between race or ethnicity and demand for health care services is a complex function of cultural, socioeconomic, and other considerations. Consequently, Freiman concludes that separate demand equations should be estimated for people in different racial or ethnic groups. To support his conclusions, Freiman presents findings from a multiple regression analysis of the 1987 National Medical Expenditure Survey where statistical tests performed indicate significant differences in the estimated coefficients of demand equations-estimated separately for non-Hispanic whites, African Americans, and Hispanics-that control for important determinants of health care use. The PARM provides insight on the proportion of patient care hours that physicians spend providing care to patients in three race/ethnic groups. These estimates, like those described for people in different age categories in the preceding section, are based on patterns of health care use, the size of the population in each demographic group, and the average amount of time physicians spend with patients per encounter. In physicians' offices and in hospital outpatient settings, the average time spent per visit can differ by patient depending on the patient's demographic characteristics and insurance status. In the other settings, however, there are insufficient data to test the hypothesis that physician time per visit is independent of patient demographics and insurance status. Note that differences in the age and sex distribution of the population, by race, contribute to differences in the proportion of patient care hours spent with patients of different races. In 2000, physicians spent approximately 69 percent of patient care hours with non-Hispanic whites, 13 percent with African Americans, and 18 percent with other minorities (Exhibits 3.3 and 3.4). Although the proportion of total patient care hours approximated the proportion of the population in each racial group, the distribution of hours varied by physician specialty. African Americans, who constituted approximately 12 percent of the U.S. population in 2000, used a disproportionately higher percentage of total patient care hours of emergency medicine physicians (38 percent), obstetrician/ gynecologists (17 percent), and pediatricians (16 percent). They received proportionately fewer hours from "other" surgical specialties (8 percent) and general surgeons (9 percent). The population in the "other" race/ethnicity category, which constituted approximately 19 percent of the total population in 2000, received a relatively larger proportion of radiology (31 percent) and pathology (29 percent) services, but a relatively smaller proportion of patient care hours from urologists (11 percent), ophthalmologists (11 percent), and general and family practitioners (13 percent). If the distribution of insurance status for non-Hispanic whites were applied to other racial minorities, the total demand for physicians in 2000 would have risen significantly (see Section 5, Scenario 5) but the percentage of patient care hours by racial group would have remained relatively unchanged. The percentage of total physician patient care hours spent with non-Hispanic whites would decline by two percentage points while the percentage spent with African Americans and other minorities would rise by one percentage point for each group. For most specialties, the change in percent of time spent with patients in each race/ethnicity group changes by less than two percentage points. The largest change is for obstetrics/gynecology services. Under this scenario, the percentage of hours spent with non-Hispanic white patients would fall by three percentage points while the percentage of hours spent with patients in the "other" category (which includes Hispanics) would rise by three percentage points. If health care utilization patterns and physician productivity patterns remain constant over time, in 2020 physicians will be spending approximately 14 percent of patient care hours with African Americans and 26 percent of hours with patients of other minority groups, again percentages roughly comparable to each group's share of the total population. Physical therapists, optometrists, and podiatrists are seen to spend a disproportionate amount of time with non-Hispanic whites relative to their share of the population (Exhibit 3.4). While the gap for African Americans is small (and non-existent in the case of podiatrists), the gap for other minority groups was large in 2000 and projected to remain so in 2020.

	Exhibit 3.3: Distribution of Total Patient Care Hours, by Patient Race: Total Active Physicians in Patient Care
	Exhibit 3.3: Distribution of Total Patient Care Hours, by Patient Race: Total Active Physicians in Patient Care (Text Only)
	Non-Hispanic White African American Other 2000 0.69 0.13 0.18 2020 0.6 0.14 0.26
	Exhibit 3.4. Estimated Percentage of Patient Care Hours, by Race of Patient
	Specialty 2000a 2020a Non-Hispanic White AfricanAmerican All Other Non-Hispanic White AfricanAmerican All Other Total Patient Care Physicians (MDs and DOs) 69 13 18 60 14 26               General Primary Care 72 13 15 63 14 24 GP & FP 78 10 13 69 11 20 General Internal Med. 72 14 14 63 15 23 Pediatrics 61 16 23 51 17 32 Medical Specialties 71 13 16 62 13 25 IM Subspecialties 71 13 16 62 13 25 Cardiovascular Diseases 73 11 15 64 12 24 Other Medical Specialties 70 13 17 60 13 26 Surgery 71 12 17 62 12 26 General Surgery 70 9 22 59 9 32 Obstetrics/Gynecology 66 17 17 57 18 25 Otolaryngology 75 11 14 67 12 21 Orthopedic Surgery 72 11 17 62 11 27 Urology 78 11 11 71 12 17 Ophthalmology 78 10 11 71 11 18 Other Surgical Specialties 73 8 19 62 8 30 Other Patient Care 64 15 21 53 15 32 Psychiatry 73 11 16 62 11 26 Anesthesiology 66 14 21 56 13 31 Emergency Medicine 47 38 16 39 39 22 Radiology 56 14 31 45 12 43 Pathology ??60 11 29 48 10 42 Other Specialties ??67 13 20 57 13 30 Non-Physician Specialties Physical Therapy 80 10 10 74 12 15 Optometry 80 10 11 73 11 16 Podiatry 78 12 10 71 14 15 Total U.S. Population 69 12 19 61 13 26
	a These forecasts from the Physician Aggregate Requirements Model assume no change over time in per capita utilization, physician productivity or mix, or the health care operating environment. Note: percentages might not sum to 100 percent due to rounding.
	3.3 Implications of the Changing Racial and Ethnic Composition of the Population for the Supply of Health Workers
	One of the five major recommendations of the Pew Health Professions Commission is to "ensure that the health profession workforce reflects the diversity of the nation's population." (O'Neil et al., 1998, p. iv). Currently, minorities are underrepresented in the physician and registered nurse workforce. The Pew Commission and numerous others argue that increasing minority representation in the health workforce is not only a commitment to diversity, but will also improve the health care delivery system. The two main arguments that diversity improves health care delivery are (1) minority health professionals express a greater propensity than do non-minority professionals to practice in underserved areas, and (2) health professionals who share the same culture and language with the patients they serve can provide more effective care (see, for example, Trevino, 1994). Much of the literature on willingness to practice in underserved areas pertains to physicians. Supply models generally do not have a race/ethnicity component. Possible reasons include data limitations and the lack of priority this topic has received. Consequently, our understanding of the relationship between supply of health workers and race/ethnicity consists of snapshots of the racial and ethnic distribution through surveys and periodic efforts to survey health workers regarding the relationship between race/ethnicity and workforce issues (e.g., workforce participation, retention, and productivity). The following are important factors and questions to consider regarding the relationship between race/ethnicity and the supply of health workers: Minorities have historically been underrepresented in higher-paying health care occupations and overrepresented in lower-paying health care occupations relative to their percentage of the U.S. population. As minorities constitute an increasing proportion of the population entering the workforce for the first time, minority representation in higher-paying health care occupations will naturally increase. As a result, the U.S. will increasingly rely on minority health workers. One area where additional research is needed is whether lifetime labor force participation patterns of health workers differ by race or ethnicity. For example, minorities have historically had a shorter life span than non-minorities, although in recent years longevity of minorities has increased to more closely resemble longevity of whites. Do differences in longevity affect retirement rates? Does race or ethnicity affect the health worker's choice of profession or medical specialty? Does race or ethnicity affect the education and training opportunities of persons desiring to enter a health care profession? Empirical research suggests that physician race and ethnicity are significant determinants of where the physician will practice, and minority physicians have a greater propensity than do non-minority physicians to practice in underserved locations. Does physician race or ethnicity affect the quality of care that patients receive? A large body of literature explores the issue of "cultural competence," which is that health professionals can provide more effective and efficient services if they are sensitive to their patients' cultural background. [8] Cross et al. (1999) define cultural competence as ? set of congruent behaviors, attitudes, and policies that come together in a system, agency, or among professionals and enable that system, agency, or those professionals to work effectively in cross-cultural situations.?It should be noted that culture is defined by more than race and ethnicity, it also encompasses economic and social factors. Often, race and ethnicity are correlated with these economic and social factors, which can obscure the relationship between health care and race or ethnicity. For example, Kington and Smith (1997) analyzed the relationship between socioeconomic status and racial and ethnic differences in the prevalence of diabetes, heart conditions, hypertension, and arthritis.They find that socioeconomic status plays a greater role in explaining racial and ethnic differences in individuals?ability to function once someone is ill, rather than explaining the differences in the probability of becoming ill. A recent literature review analyzed over 120 works in the field of cultural competence in health care to develop a measurement profile for cultural competence in health care delivery settings (The Lewin Group, 2001). Several articles reviewed discuss the need to emphasize the value of diversity and the importance of involving people from diverse backgrounds in the decision-making process of what care to provide to underrepresented groups. Much of the literature discusses the importance of educating health workers to be sensitive to differences in the needs of their patients resulting from differences in culture. Brown and Nichols-English (1999) discuss the implications of patient diversity for pharmacists. People of different cultures-which they broadly defined by race and ethnicity, language, socioeconomic group, family structure, and geographic location-have different perceptions, on average, of health care issues. Their perceptions might differ in the following: "(1) [the constitution of] disease and its causation; (2) appropriate health-care-seeking behavior; (3) the quality and usefulness of medical encounters; (4) effective approaches to healing, including both conventional and alternative practices; and (5) the role of family in health care (p. 61)." Brown and Nichols-English discuss the importance of educating pharmacists on providing culturally competent care to reduce drug-related problems-e.g., noncompliance, adverse effects, and sub-optimal dosing.
	3.3.1 Physician Supply
	Relative to the overall population, minorities are underrepresented in the physician workforce for all races and Hispanic ethnicity with the exception of the population of Asian descent. Exhibit 3.5 shows the distribution of the physician workforce by race and ethnicity in 1999. For those physicians whose race and ethnicity is recorded in the AMA master file, 75.4 percent are non-Hispanic white, 3.6 percent are African American, 4.9 percent are Hispanic, 12.6 percent are Asian, 0.1 percent are American Indian or Alaskan Native, and the remaining 3.5 percent are of various other races.
	Exhibit 3.5. Race Distribution of the Physician Workforce, 1999
	Exhibit 3.5. Race Distribution of the Physician Workforce, 1999 (Text Only)
	Race Percentage White 75.4% Black 3.6% Hispanic 4.9% Asian 12.6% Other 3.5%
	Source: American Medical Association, Physician Characteristics and Distribution in the U.S., 2001-2002 Edition.

	The racial and ethnic composition of the physician workforce, however, varies substantially by specialty (Exhibit 3.6). The percent non-Hispanic white ranges from a high of 91.1 percent in aerospace medicine to a low of 65.2 percent in physical medicine and rehabilitation. Specialties with the highest representation of physicians of Asian descent are physical medicine and rehabilitation (20.6 percent), internal medicine (17.9 percent) and radiation oncology (17.4 percent). African Americans have the highest representation in general preventive medicine (6.3 percent), obstetrics/gynecology (6.2 percent) and pediatrics (4.8 percent). Hispanics have the highest representation in general practice (7.9 percent), child psychiatry (7.0 percent) and pediatrics (6.7 percent).
	A visual inspection of the specialties where physicians spend relatively more (less) time with African American and other minority patients (Exhibit 3.4) finds that these specialties tend to have higher (lower) minority representation in the physician workforce. The three specialties, for example, shown in Exhibit 3.4 to have spent the greatest percentage of time with African American patients in the year 2000 were emergency medicine, obstetrics/gynecology, and pediatrics; from Exhibit 3.6, we see that each of these specialties had in 1999 an above-average representation of African American physicians compared to the workforce at large (4.1, 6.2, and 4.8 percent respectively, compared to an overall average of 3.6 percent for all specialties combined). The two specialties shown in Exhibit 3.4 to have spent the lowest percentage of time with African American patients in the year 2000 were general surgery and other surgical specialties, groups characterized in Exhibit 3.6 by a below-average representation of African Americans. Similar observations apply, with some exceptions, to Hispanics and other minorities. The exceptions are as follows: (a) radiologists spent a large percentage of time with "other minority" patients (31 percent) despite the fact that other minorities constituted a distinctly below-average percentage of the radiologist workforce (12.7 percent as against an overall average of 21 percent), and (b) cardiologists spent a low percentage of time with other minority patients (15 percent) despite the fact that other minorities constituted an above-average percentage of the cardiologist workforce (26.3 percent compared to 21 percent for all specialties combined). Advocates for increased representation of minorities in the physician workforce cite both equity and efficiency reasons. One equity issue cited is providing greater access to care for minority populations who are disproportionately in designated physician shortage areas. Defining a "physician shortage area" as an area with fewer than 30 office-based primary care physicians per 100,000 population, Komaromy et al. (1996) found that 57 percent of poor areas with a high percentage of African American and Latino residents could be classified as physician shortage areas. In comparison, Komaromy et al. found that only 13 percent of poor areas with a high percentage of non-Hispanic white residents could be classified as physician shortage areas. Intuitively, one might expect that poorer urban neighborhoods might naturally have fewer physicians per population. Komaromy et al. found, however, a stronger correlation between the physician supply and the proportion of residents in the community who are African American or Hispanic residents than the correlation between physician supply and an area's average income level.

 

	Exhibit 3.6. Percent Distribution of Physicians by Race and Ethnicity, in 1999
	Specialty Non-Hispanic White African American Hispanic Asian Other American Indian/ Alaskan Native Total MDs 75.4 3.6 4.9 12.6 3.5 0.1 Aerospace Medicine 91.1 2.1 3.4 2.1 1.3 0.0 Allergy & Immunology 79.0 1.4 3.7 12.2 3.6 0.1 Anesthesiology 71.5 3.4 4.2 16.9 4.1 0.1 Cardiovascular Disease 71.1 2.4 4.9 15.2 6.2 0.1 Child Psychiatry 73.5 4.8 7.0 10.2 4.3 0.2 Colon/Rectal Surgery 81.6 1.8 5.1 9.8 1.7 0.0 Dermatology 87.4 2.5 2.7 5.9 1.4 0.0 Diagnostic Radiology 80.2 2.0 3.5 11.5 2.7 0.1 Emergency Medicine 82.4 4.1 4.2 7.3 1.8 0.1 Family Practice 79.2 4.1 5.4 8.8 2.3 0.2 Forensic Pathology 82.2 3.7 4.7 8.3 0.7 0.3 Gastroenterology 71.5 3.1 4.8 14.8 5.8 0.0 General Practice 75.5 2.2 7.9 13.6 0.7 0.0 General Preventive Med. 82.1 6.3 3.5 6.6 1.4 0.1 General Surgery 78.3 3.4 4.6 10.9 2.7 0.1 Internal Medicine 67.0 4.1 5.1 17.9 5.8 0.1 Medical Genetics 84.6 1.9 3.4 8.0 2.2 0.0 Neurology 72.8 1.9 5.2 14.1 5.9 0.0 Neurological Surgery 82.7 2.5 3.9 7.8 3.1 0.1 Nuclear medicine 71.0 2.0 5.7 16.8 4.4 0.0 Obstetrics/Gynecology 77.2 6.2 5.3 9.3 2.0 0.1 Occupational Med. 88.6 2.9 3.0 4.6 0.9 0.1 Ophthalmology 84.6 2.2 2.9 7.9 2.4 0.1 Orthopedic Surgery 88.7 2.4 2.4 4.7 1.7 0.1 Otolaryngology 84.3 2.0 3.2 8.7 1.7 0.0 Pathology-Anat/Clin 74.5 1.9 4.9 15.5 3.3 0.0 Pediatrics 68.6 4.8 6.7 15.7 4.0 0.1 Pediatric Cardiology 75.7 2.0 5.6 11.5 5.0 0.1 Physical Med/Rehab 65.2 4.4 6.2 20.6 3.7 0.1 Plastic Surgery 84.8 1.8 3.6 7.5 2.3 0.0 Psychiatry 75.0 3.2 5.6 12.7 3.4 0.1 Pulmonary Diseases 75.7 2.6 5.2 12.3 4.2 0.1 Radiology 85.7 1.4 2.2 8.8 1.7 0.0 Radiation Oncology 73.5 2.7 3.2 17.4 3.2 0.0 Thoracic Surgery 67.2 4.2 5.6 10.5 12.2 0.3 Urological Surgery 82.7 2.8 3.5 8.7 2.2 0.1 Other 88.2 2.0 3.4 5.5 0.9 0.0
	Source: American Medical Association, Physician Characteristics and Distribution in the U.S., 2001-2002 Edition.

	The Komaromy et al. study found that of many possible characteristics of a physician, the best predictor for whether the physician cared for a high percentage of African American patients was whether the physician was African American. After controlling for the proportion of African American residents in the community, this analysis indicated that the proportion of African American patients cared for by African American physicians was 25 percentage points higher than the average proportion of African American patients cared for by physicians of other races.
	Other variables, such as the ranking of the physician's medical school, experience, and type of hospital had insignificant effects. The authors suggest that the personal choice of the physician is the most likely explanation for the phenomenon that African American physicians are more likely than non-Hispanic white physicians to treat African American patients. Given that the ranking of the physician's medical school is not significant in predicting the race of the physician's patients, the authors conclude that top African American medical school graduates are themselves choosing to practice in poorer, predominantly minority areas A study by Moy and Bartman (1995) found that minority patients were more than 4 times as likely as non-Hispanic white patients to receive care from minority physicians. Moy and Bartman note that any solution that attempts to increase the proportion of minority physicians must also take into account the financial hardships they face. On average, minority physicians tend to treat lower-paying uninsured and Medicaid patients. Moy and Bartman estimate Medicaid fees for physician services as averaging only 47 percent of private insurance fees. Because up to 29 percent of low-income patients are receiving care from minority physicians, these physicians must bear a disproportionately higher share of the financial burden associated with poorer patients. Medicaid insured 45 percent of the patients seen by African American physicians and only 18 percent of patients seen by non-Hispanic white physicians. Hispanic physicians cared for more uninsured patients than physicians of other ethnic groups. On average, 9 percent of their patients were uninsured compared to 6 percent for non-Hispanic white physicians. Physicians whose clientele is composed of a high percentage of Medicaid and uninsured patients may also have a more difficult time securing managed care contracts. Bindman et al. (1998) studied the frequency of denials or terminations of managed care contracts experienced by primary care physicians. They found that physicians with higher proportions of uninsured patients were 4 times more likely to have a contract terminated or denied. There was also a significant positive correlation between the number of uninsured patients a physician saw and the frequency of denials from managed care contracts for these physicians. Latino physicians had significantly lower odds of having more than 10 percent of their patients enrolled in a managed care plan: 23 percent of Latino physicians in group practice are in no way affiliated with an HMO. One reason for the imbalance, noted by Mackenzie et al. (1999), might be that solo practices are associated with lower levels of participation in managed care, and minority physicians tend to have solo practice settings. MacKenzie et al., through a survey of physicians who tended to treat managed care patients, found that 56 percent claimed they had difficulty referring patients of varied ethnic backgrounds to specialists who met those patients' cultural needs. The author expresses guarded optimism that as the idea of cultural competency within managed care gains momentum, managed care organizations will become increasingly aware of the importance of 'ethnic matching'. As a result, they may attempt to recruit ethnic minority physicians as a way to attract and retain ethnic minority members.

	In 2000, minorities constituted 27 percent of the population age 18-34-the age group that reflects the population entering the workforce. By 2020, minorities will constitute approximately 45 percent of the age 18-34 population. An estimated 15 percent of the population in this age group will be African American, and 30 percent will be Hispanic or a non-African American minority. As minorities constitute a larger portion of the population from which new health workers are drawn, minority representation in the physician workforce will naturally rise. Also, as noted by Libby, Zhou, and Kindig (1997), organizations such as the Bureau of Health Professions, the Institute of Medicine, the Association of American Medical Colleges, and others have made racial/ethnic equity in the physician workforce a high priority. As shown by Libby et al., however, racial parity in the physician workforce will likely not occur in the next few decades, although some gains in parity will be made. For five race/ethnicity groups, these authors forecast the number of physicians per 100,000 population of the same race or ethnicity as the physician. Their projection model constrains the race-specific physician-to-population ratios to converge over time to an equilibrium of 218 physicians per 100,000 population by adjusting the racial composition of first-year graduate medical education cohorts. The soonest that racial parity is reached, given projected demographics, is around 2040. In summary, the literature and changing demographics suggests that increasing minority representation in the physician workforce will improve access to care for minority and vulnerable populations. Minorities face financial obstacles to become physicians, and once they become physicians may face greater financial obstacles than non-minority physicians because of practice location or other factors. Increased racial/ethnic diversity of the U.S. population over the next few decades will naturally increase minority representation in the physician workforce.
	3.3.2 Nurse Supply
	Estimates from the 2000 Sample Survey of Registered Nurses (HRSA, 2001) indicate that approximately 86.6 percent of RNs are non-Hispanic white, 4.9 percent are non-Hispanic African American, 3.5 percent are Asian; 2 percent are Hispanic; 0.5 percent are American Indian or Alaskan Native, 0.2 percent are Native Hawaiian or Pacific Islander, and 1.2 percent are of two or more racial backgrounds (see Exhibit 3.7). Among minorities, Hispanics and African Americans are underrepresented in the registered nurse workforce relative to their proportion in the overall population.

	Exhibit 3.7 Racial/Ethnic Distribution of the Registered Nurse Workforce in 2000
	Exhibit 3.7 Racial/Ethnic Distribution of the Registered Nurse Workforce in 2000 (Text Only)
	Race/ethnicity % RN Population % U.S. Population Caucasian (non-Hispanic) 86.6% 69.1% African American(non-Hispanic) 4.9% 12.1% Hispanic 2.0% 12.5% American Indian,Alaskan Native,Hawaiian/Pacific Islander 0.7% 0.8% Asian 3.5% 3.6% Two or more races 1.2% 1.6%
	Source: 2000 Sample Survey of RNs (HRSA, 2001).

	The literature on the relationship between race or ethnicity and the supply of nurses is substantially smaller than the corresponding literature for physicians. Sechrist, Lewis, and Rutledge (1999) report that the nurse workforce in California is becoming more ethnically diverse. Although minorities are underrepresented in the current nurse workforce in California, the racial and ethnic mix of nursing school entrants more closely parallels the diversity of California’s population. The authors report that minority students, however, are less likely to graduate from nursing programs than their non-Hispanic white counterparts.
	The authors make several recommendations to improve ethnic diversity of the nurse workforce including outreach efforts to increase the number of minorities in nursing programs. They cite an unpublished study by Martin-Holland et al. that looks at strategies to improve ethnic diversity in the nurse workforce. Specifically, the study looks at (1) strategies that have been successful in recruiting and retaining ethnically diverse students in nursing programs, (2) barriers to nursing program success for ethnically diverse students, and (3) activities incorporated into nursing programs to improve cultural sensitivity of nursing school graduates. In 2000, approximately 61 percent of the female population age 18-34—the main source of new nurses—was non-Hispanic white. By 2020, the percentage will have decreased. Only half of all women age 18-34 will be non-Hispanic white; African Americans and all other minorities (including white Hispanics) will constitute 16 percent and 33 percent, respectively, of the female population age 18-34. As minorities constitute a growing proportion of the female population in this group, minority representation in the nurse workforce will naturally rise. Furthermore, the growing nurse shortage in the U.S. has encouraged some employers to recruit foreign nurses. Recruiting foreign nurses will increase the diversity of the nurse workforce; however, many of the countries exporting nurses to the U.S. may themselves in turn face an inadequate supply of nurses.[9] Wall Street Journal article: Shortage of Nurses Hits Hardest Where They Are Needed the Most: Nurse Shortage Shows How Labor Markets Go Global (Zachary, 2001).
	GEOGRAPHIC LOCATION OF THE POPULATION
	Discussion of the adequacy of the health care workforce is often framed in the context of a maldistribution of workers. An inadequate supply of health workers is often a local or regional phenomenon, Major Findings: Geographic variation in population growth rates and determinants of health worker demand and supply highlight the importance of developing forecasting models that can make State-level and sub-State level forecasts. Although an increasing proportion of the U.S. population resides in urban areas, a substantial proportion of the population will continue to reside in rural areas. Many of these rural areas are currently designated as physician shortage areas. Pockets of urban areas will continue to have a high concentration of minorities. Many of these areas are currently designated as physician shortage areas. frequently accompanied by surpluses elsewhere. Consequently, national forecasts of supply and demand can mask inadequacies of supply at the local level. Trends in geographic location of the population that have important implications for the future health care workforce include the following. First, there is substantial variation in population growth and other factors that affect the supply of and demand for health professionals. This phenomenon highlights the importance of models that can forecast at the State and local level. Second, a significant proportion of the population will continue to reside in rural areas and have less access to health care services than the population residing in urban areas. Third, some urban areas will continue to have a high concentration of minorities. These areas are often characterized as having fewer economic resources per capita, greater health care needs, and less access to health care services than surrounding areas.
	4.1 Population Projections and Regional Growth Patterns
	According to the U.S. Census Bureau (Campbell, 1997), all regions of the country will grow over the next 25 years, with the West and the South growing at the fastest rate (Exhibit 4.1). As the population continues to rapidly grow in these regions, the demands for health care will also increase.

	Exhibit 4.1 Population Projections by Region
	Region Projections by Year (thousands of resident population) 1995 2000 2005 2010 2015 2020 2025 Northeast 51,466 52,107 52,767 53,692 54,836 56,103 57,392 Midwest 61,804 63,502 64,825 65,915 67,024 68,114 69,109 South 91,890 97,613 102,788 107,597 112,384 117,060 121,448 West 57,596 61,413 65,603 70,512 75,889 81,465 87,101 Total 262,755 274,634 285,981 297,716 310,133 322,742 335,050
	Source: United States Census Bureau (Campbell, 1997).

	The uneven regional growth of the population has both short-term and long-term ramifications for the health workforce. Regions of the country that experience rapid growth in population could experience temporary shortages of some health professionals, such as physicians, who might be less mobile than the population at large. Efforts by some localities to recruit specific growth industries-e.g., high-tech industries-without a balanced approach to recruit health professionals could cause a short-term strain on the local health care infrastructure. Areas of the United States that are already experiencing physician shortages and that are high-growth areas might see more severe short-term inadequacies in the health workforce. For example, the Census Bureau estimates that Texas will be one of the fastest growing States over the next 20 years. However, according to the Bureau of Primary Health Care, Texas currently has one of the highest number of physician shortage areas in the country, understandable in view of its size. Not only does this trend appear in Texas, but many smaller southern States also face a combination of high growth and a large number of shortage areas. Regional differences in physicians per population and nurses per population do not necessarily reflect inadequacies in the health care workforce. As discussed previously, demand for health care services is highly correlated with the age distribution of the population, and there is substantial geographic variation in the age distribution of the population. For example, the proportion of the population age 65 and older is much greater in Florida (18), West Virginia (17) and North Dakota (15) than it is in Alaska (5), Utah (8) and Colorado (9). In addition, there exists substantial variation in other determinants of demand for health care services such as the characteristics of the health care operating environment, economic conditions, and lifestyle. Douglass (1995) projected the future supply of family physicians on a State-by-State basis and found substantial regional variation in physician supply and needs. One implication of the uneven population growth and geographic variation in the determinants of supply and demand is the need to develop forecasting models that can forecast at the State or sub-State level. The NDM forecasts demand for nurses at the State level. Preliminary demand forecasts compared to current and future supply forecasts show substantial variation across States in the adequacy of the nurse workforce-both now and in the future (Dall and Hogan, 2002).
	4.2 Evolving Trends in Urbanization
	Although the proportion of the U.S. population living in metropolitan areas will continue to grow, a large proportion of the population will continue to live in rural areas. A substantial body of literature describes the inadequacies of the physician workforce in rural areas, and over 65 of the Health Professional Shortage Areas (HPSAs) are in rural areas. Between 1990 and 2000, the population in metropolitan areas increased by nearly 14 percent, whereas the population in non-metropolitan areas grew by only 10 percent (Exhibit 4.2). One reason for this phenomenon is a matter of classifications: geographic regions formerly designated as rural areas are becoming more metropolitan and were re-designated as metropolitan areas. Another reason is immigration: immigrants disproportionately settle in metropolitan areas. A third reason is migration from rural to urban areas, although this effect has been small. The Census Bureau (March 2001) reports that net migration out of rural areas totaled only 137,000 between 1998 and 2000. The "metropolitanization" of the country could help alleviate the problems of an inadequate supply of physicians in some rural locations as the population in these areas increases above the threshold required to support a more comprehensive health workforce.

	Exhibit 4.2 Population Growth by Metropolitan Status and Size
	Population Size Population Percent Change 1990-2000 2000 share of total   April 1, 1990 April 1, 2000 United States 248,709,873 281,421,906 13.2 100.0 Total Metropolitan 198,402,980 225,981,676 13.9 80.3 5 million or greater 75,874,152 84,064,274 10.8 29.9 2 – 5 million 33,717,876 40,398,283 19.8 14.4 1- 2 million 31,483,749 37,055,342 17.7 13.2 250,000 – 1 million 39,871,391 45,076,105 13.1 16.0 250,000 or fewer 17,455,812 19,387,675 11.1 6.9 Non-Metropolitan 50,306,893 55,440,227 10.2 19.7
	Source: United States Census Bureau.

	Substantial proportion of the population will continue to reside in rural areas during the foreseeable future. When modeling the supply of health professionals in rural and underserved areas, analysts might consider the following obstacles to increasing physician supply in these shortage areas, as reported in the literature. Connor, Hillson and Krawelski (1995) suggest that physicians locate in areas with other physicians in order to benefit from the professional synergism that develops when there is an established population of physicians. Similarly, Brasure et al. (1999) found a general aversion to rural practice may exist among urban professionals, but there is less resistance to enter an underserved market once at least one health provider has settled there. Efforts to model the supply of physicians in underserved areas might identify "forerunner" specialties and analyze patterns of physician location. Olchanski et al. (1998) found that the average age of physicians in rural areas of Virginia is increasing, raising concerns that physician shortages in these areas will be exacerbated when these physicians retire. Furthermore, he speculates that this phenomenon could be applicable to other parts of rural America. Rabinowitz et al. (1999), in a study of rural physicians in Pennsylvania, found that one of the most critical factors in determining whether a physician will practice in a rural environment is the extent of the physician's rural background. Models of physician supply might incorporate an urban/rural dimension that takes into account the propensity of physicians to practice in physician shortage areas based on the background and demographic characteristics of medical students and the existing physician workforce. A disincentive to physicians choosing to practice in rural settings is lower earnings potential. For heavily-indebted physicians exiting medical school, practicing in suburban areas where there is greater economic activity can be more enticing than practicing in a rural area. Government and private organizations have implemented various programs and grants to encourage physicians to practice in underserved, rural areas. For example, the State of Illinois, along with the University of Illinois College of Medicine at Rockford, has implemented a program designed to improve the supply of physicians to these areas. According to Stearns et al. (2000), this program has been reasonably successful, with 69 percent of the graduates choosing to enter rural practices. Efforts to model physician supply might incorporate estimates of the impact of programs that try to influence where physicians will practice. Similarly, some States are offering grants to people in nursing programs who agree to work in rural or underserved areas for a specific length of time following graduation. Some researchers have argued that international medical graduates (IMGs) can be used to augment the physician workforce in underserved areas. Mick et al. (2000, 1999) have shown that the IMGs are more likely than U.S. medical graduates to locate in rural areas with high rates of infant mortality, fewer per capita economic resources, a high proportion of minorities, a disproportionate number of elderly, and low physician-to-population ratios. Baer et al. (1999) found that IMGs were also fulfilling an important role in community health centers. These centers tend to be located in physician shortage areas, so these researchers suggest that the role of IMGs is indispensable in the rural setting. As hospitals in rural areas close, the authors assert that community health center clinics are the most effective way for underserved populations to receive the health care they require and that IMGs help fill a 'safety net' role. Not all researchers agree that IMGs help alleviate physician shortages in underserved areas. A study conducted by Politzer, Cultice, and Meltzer (1998) found that the geographic distribution of physicians has become less even. The study also argued that IMGs, rather than helping to mitigate this trend, had in fact contributed to its severity. The authors state that the majority of IMGs choose not to work in areas with a physician shortage, and that the contributions others note are overstated.

	4.3 Urban Demography and the Effects on Physician Locations
	Pockets of the population will continue to contain high concentrations of minorities. These pockets, generally located in urban areas, are often characterized by lower average levels of economic resources, greater average health care needs, and less access to health care services. COGME (1998) reports that although there appears to be an oversupply of physicians, most of the oversupply is located in affluent urban and suburban areas. Additionally, specialists are especially prone to locating in more affluent areas. The traditionally poor areas of the city exhibit a unique need, as they are often demographically independent from the more affluent areas in the same region. One of the most sensitive populations is the immigrant population, especially those with little or no English proficiency. Members of this population tend to locate in areas that traditionally consist of low-income households and are more likely to live in cities than non-metro areas. According to the 2000 census, 5.1 percent of foreigners live in rural areas, compared to 20.7 percent of native-born people. This means that as immigration increases, there may be greater pressure placed on urban community hospitals, which typically serve more non-English speaking people (Gaskin and Hadley, 1999). According to Gaskin and Hadley, these hospitals face a higher level of physician and health care professional shortages, thus degrading the level of care provided to the underserved population. As immigration increases in the near future, this strain placed on the community hospitals may increase. In addition to the use of IMGs in rural areas, Mick has suggested that they may help relieve shortages in the urban areas as well. According to his study, IMGs tend to locate in less affluent areas within a city and are willing to work for a lower salary. Additionally, as discussed previously, some policy makers advocate increasing the efforts made towards recruiting minorities into the health care professions. They claim that these individuals may be willing to work in shortage areas, as well as being able to overcome some of the language barriers that exist in some of these areas (Trevino 1994, Komarmony et al., 1996).

	MODELING THE IMPACT OF CHANGING DEMOGRAPHICS ON THE FUTURE DEMAND FOR HEALTH PROFESSIONALS
	Efforts to model the impact of changing demographics on the demand for and supply of health professionals incorporate many of the demographics trends discussed above as well as trends in economics, technology, the education system, regulation and legislative activities, the health care operating environment, and the ability to substitute between health professionals. Recent modeling efforts differ in level of sophistication, factors used to forecast future supply and demand, and assumptions made by analysts. A consensus exists that the supply of physicians and nurses can be predicted with an adequate degree of accuracy even 10 or 20 years into the future (see, for example, Tarlov [1995] and Prescott [2000]). Previous efforts to model the requirements for health workers, on the other hand, have met with mixed success and often with controversy. As discussed above, efforts over the past two decades to model requirements show there is little consensus on how best to define requirements, the relationship between requirements and its determinants, the future values of many of these determinants, and forecasters' assumptions. There is often disagreement regarding how requirements should be defined. For example, should requirements be defined by an assessment of the population's needs? Should requirements be based on demand and, if so, are current levels of employment accurate measures of demand? Should requirements be defined by benchmarking? For example, one could compare physician staffing levels to a level determined to be "efficient" (e.g., HMO staffing patterns). Alternatively, one could compare physician-per-population levels in the U.S. to levels in other countries. Or, should requirements be defined as some combination of demand, needs, and benchmarking? Despite these concerns and disagreements, supply and demand models are important tools to help analysts and policy makers understand the implications of trends and policies. This section contains a brief description of two requirements forecasting models recently updated by BHPr-the Physician Aggregate Requirements Model (PARM) and the Nursing Demand Model (NDM)-and presents preliminary forecasts of the impact of changing demographics and other user-defined scenarios on requirements for the health professions in these two models. Both models define requirements as the number of health workers that the U.S. is likely to demand based on population needs and economic considerations. Demographics, especially the growth in size of the elderly population, are the driving force behind most projections of future workforce requirements. Future demographics can be extrapolated with some degree of accuracy based on historical patterns of fertility rates, mortality rates and migration. The Census Bureau publishes its middle series projections that extrapolates future population levels based on expected fertility, mortality, and migration patterns. The Census Bureau last updated the series in 1996, and the middle series under-predicted the size of the 2000 population by approximately 6.8 million individuals (or 2.4 percent of the total population). The population projections used in the PARM and NDM are based on the Census Bureau's middle series projections, but incorporate adjustments based on recently released 2000 census data.

	5.1 Physician Aggregate Requirements Model
	The PARM combines projections of the future demand for health care services, by medical specialty and setting, with estimates of physician productivity to forecast future requirements. Exhibit 5.1 provides an overview of this process. For a more thorough description of the model and its capabilities see PARM User Guide and Technical Report (Dall, 2002). To calculate future demand for health care services, the PARM first combines population projections (Exhibit 5.2) by six age groups, three race/ethnicity groups, and sex (Box 1 of Exhibit 5.1) with estimates of the proportion of the population in each of three insurance categories (Box 2) to divide the population into 108 categories (Box 3). The six age categories are 0-17, 18-34, 35-54, 55-64, 65-74, and 75 years and older. The three race categories are non-Hispanic white, African American (Hispanic and non-Hispanic), and other (including white Hispanic). The three insurance categories are (1) the insured who receive services in a fee-for-service arrangement, (2) people enrolled in a health maintenance organization (HMO), and (3) the uninsured. The PARM contains 22 categories of health professionals providing patient care. These categories consist of 19 physician specialties and three non-physician specialties (i.e., physical therapy, podiatry, and optometry). The process to forecast requirements is similar for both physicians and these three non-physician specialties, although the data sources differ. The workload measures used in the PARM are physician-patient encounters in each of five settings: (1) doctors' offices, (2) hospital outpatient clinics and emergency departments, (3) hospital inpatient (hospital rounds), (4) hospital inpatient (surgery), and (5) other settings (e.g., nursing homes and home health). The PARM multiplies the number of people in each population category by its corresponding estimate of per capita physician-patient encounters (Box 4) to estimate total demand for physician services as measured by physician-patient encounters (Box 5). Estimates of total encounters in each setting (Box 5), multiplied by the average minutes physicians spend per encounter (Box 6), creates an estimate of total physician minutes required to provide patient care (Box 7). Note that the minutes per encounter include an adjustment for indirect patient care to capture time spent on tasks such as completing paperwork and reviewing patient histories. Total required minutes (Box 7), divided by estimates of total annual patient care minutes per physician in each specialty (Box 8), creates forecasts of total physician requirements for patient care activities (Box 9). The data on physician-patient encounters and physician productivity come from the AMA annual survey and thus only include MDs. Consequently, an adjustment is made to the physician requirement counts to include DOs (Box 10). Data on the number of DOs in 1999, by specialty, come from the American Osteopathic Association. These numbers are inflated, using recent growth rates by DO specialty, to update the numbers to the base year of 2000. In addition, requirements for physicians in non-patient care activities (e.g., administration, teaching, and research) are calculated as a fixed percentage of physicians in patient care. Calibration adjustments are made to equate base year forecasts of actual physician supply with base year estimates of total requirements (Box 11), and this produces the refined forecasts of requirements for the 22 original specialties plus a category for physicians in nonpatient care activities. The base year for total MD counts is 2000.[10] The base year counts of MDs come from the AMA's Physician Characteristics and Distribution in the US: 2002-2003 Edition. Active MDs whose specialty is unknown are distributed across the other specialties based on those specialties' proportion of total active physicians. The shaded boxes (i.e., boxes 2, 4, and 6) indicate areas of the PARM where the user can easily change the forecasting assumptions.

	Exhibit 5.1 PARM Structure

 

	Exhibit 5.2 U.S. Population Forecasts (in thousands)
	Race Sex Age Year 1999 2000 2005 2010 2015 2020 Non-Hispanic White Men 0-17 22,737 22,628 22,042 21,315 21,067 21,143 18-34 21,373 21,223 21,069 21,375 21,641 21,174 35-54 29,670 29,974 29,654 27,994 25,887 24,596 55-64 9,027 9,231 11,341 13,104 14,371 14,675 65-74 6,871 6,846 6,894 7,854 9,817 11,599 75+ 5,141 5,255 5,686 6,019 6,367 7,341 Men Total 94,818 95,158 96,685 97,661 99,150 100,528 Women 0-17 21,501 21,399 20,851 20,152 19,902 19,965 18-34 21,003 20,842 20,597 20,842 21,116 20,672 35-54 29,896 30,214 29,996 28,385 26,258 24,909 55-64 9,594 9,796 11,931 13,730 14,967 15,241 65-74 8,233 8,132 7,919 8,789 10,750 12,485 75+ 8,887 9,011 9,352 9,318 9,395 10,184 Women Total 99,113 99,395 100,646 101,216 102,389 103,456 Non-Hispanic White Total 193,931 194,553 197,332 198,877 201,539 203,984 African American Men 0-17 5,483 5,532 5,799 5,987 6,282 6,619 18-34 4,305 4,319 4,474 4,765 5,052 5,276 35-54 4,374 4,483 4,768 4,747 4,697 4,722 55-64 1,029 1,057 1,317 1,637 1,969 2,149 65-74 660 666 707 864 1,096 1,404 75+ 400 408 450 473 517 594 Men Total 16,252 16,465 17,515 18,472 19,613 20,763 Women 0-17 5,310 5,354 5,593 5,754 6,017 6,322 18-34 4,643 4,653 4,800 5,053 5,338 5,566 35-54 4,999 5,125 5,477 5,626 5,582 5,602 55-64 1,277 1,313 1,634 2,102 2,512 2,735 65-74 937 947 1,009 1,136 1,423 1,802 75+ 791 802 864 876 949 1,071 Women Total 17,957 18,193 19,376 20,547 21,821 23,096 African American Total 34,209 34,658 36,892 39,020 41,434 43,859 Other (including Hispanic White) Men 0-17 8,676 8,899 10,050 11,092 12,317 13,752 18-34 8,340 8,453 9,160 9,063 10,324 11,398 35-54 6,221 6,488 7,627 10,078 10,741 11,450 55-64 1,298 1,357 1,786 2,408 3,051 3,710 65-74 761 791 945 1,224 1,608 2,111 75+ 418 443 585 764 963 1,226 Men Total 25,713 26,431 30,153 34,630 39,004 43,647 Women 0-17 8,269 8,482 9,577 10,570 11,739 13,100 18-34 7,390 7,546 8,439 9,077 10,303 11,348 35-54 6,288 6,543 7,618 9,524 10,354 11,259 55-64 1,452 1,520 2,015 2,709 3,339 3,943 65-74 976 1,009 1,176 1,486 1,941 2,495 75+ 642 681 901 1,182 1,455 1,810 Women Total 25,017 25,780 29,725 34,549 39,130 43,955 Other Total 50,730 52,211 59,877 69,179 78,134 87,602 Total U.S. Population 278,870 281,422 294,100 307,075 321,107 335,444
	Source: Modified version of Census Bureau middle series projections.

	The base year for the PARM is 2000; however, data from 1996 to 2000 are pooled from some health care use databases to increase sample size. Data from the 1999 National Health Interview Survey (NHIS) are used to estimate the proportion of people in each demographic category among three possible insurance status groups.

	5.1.1 Modeling Physician Requirements
	To estimate per capita demand for physician services from each of the 108 population groups in the PARM, we first estimated the total amount of care that physicians in each specialty provide in each setting. We estimated these totals using AMA estimates for 1999 of the total number of MDs in each medical specialty primarily engaged in patient care, and data from the 1998 and 1999 AMA physician surveys that asked respondents the average number of weeks worked per year and average encounters (i.e., visits or surgical procedures) per week with patients. These data come from the 1999-2000 and 2000-2002 editions of the Physician Socioeconomic Statistics. Published statistics from the 1998 and 1999 surveys were averaged because sample sizes for some specialties are relatively small. We used the following databases to determine the distribution of total patient-physician encounters across the 108 population subgroups: The 1997, 1998, and 1999 National Ambulatory Medical Care Survey (NAMCS) databases were pooled to analyze patient-physician encounters in physicians' offices. The 1997, 1998 and 1999 National Hospital Ambulatory Care Survey (NHAMCS) databases were pooled to analyze patient-physician encounters in hospital outpatient and emergency department settings. The 1997 and 1998 Health Care Cost and Utilization Project (HCUP) databases were pooled to analyze patient-physician encounters in hospital inpatient settings. The 1996 and 1998 National Home and Hospice Care Survey (NHHCS) databases were pooled to analyze patient-physician encounters in patients' homes. As illustrated in Exhibit 5.1, we combine information on per capita demand for physician services obtained from an analysis of these databases with population forecasts and estimates of annual physician time spent in patient care to forecast future requirements for physicians. Below we present forecasts of physical requirements under five scenarios. We selected these scenarios based on policies being advocated in the political arena and scenarios looked at in previous modeling efforts. In all of these scenarios, changing demographics-and in particular the aging of the population-are a major determinant of the projected increase in physician requirements between 2000 and 2020. Comparing the forecasts from a particular scenario to the forecasts from scenario 1 (which represents the status quo) indicates the impact upon physician requirements attributed to changing demographics and/or changes in forecasting assumptions. Scenario 1, the status quo, assumes that patterns of health care use, insurance distribution, physician staffing, and physician productivity remain constant over time similar to the patterns that existed in the late 1990s.[11] As discussed above, the PARM assumes that an adequate supply of physicians existed in the base year (i.e., 2000). An over (or under) supply of physicians in the base year will result in an over (or under) estimate of requirements in future years. Patterns of health care use cover the period 1996 to 1999. Under this scenario, the number of physicians would increase from approximately 781,300 in 2000 to 1,038,200 in 2020, a 33 percent increase (Exhibits 5.3 and 5.4). At the same time, the U.S. population would increase by 19 percent, so that the ratio of physician per population would rise from 2.8 per thousand population in 2000 to 3.1 per thousand population in 2020. Medical specialties experiencing the largest percentage increases in demand between 2000 and 2020 are cardiovascular diseases (52 percent), radiology (51 percent), pathology (44 percent) and various surgical specialties (44 percent). Medical specialties experiencing the smallest percentage increases in demand are pediatrics (11 percent), obstetrics/gynecology (14 percent) and psychiatry (22 percent).

	Exhibit 5.3 Impact of Changing Demographics on Requirements for Physicians: Status Quo Scenario
	Medical Specialty 2000 2005 2010 2015 2020 % Change 2000 to 2020 Total Physicians (MDs and DOs) 781,282 831,447 891,687 959,996 1,038,234 33 Total Patient Care Physicians 733,342 780,266 836,594 900,574 973,840 33   General Primary Care 268,710 283,632 300,651 320,992 344,907 28 GP & FP 109,571 115,583 122,512 130,358 139,252 27 General Internal Med. 106,411 114,197 123,645 134,406 146,885 38 Pediatrics 52,728 53,852 54,494 56,228 58,770 11 Medical Specialties 96,926 104,145 113,200 123,560 135,331 40 IM Subspecialties 40,205 43,336 47,301 51,841 56,955 42 Cardiovascular Diseases 20,828 22,675 25,143 28,172 31,690 52 Other Medical Specialties 35,893 38,133 40,756 43,548 46,687 30 Surgery 161,160 171,133 183,519 197,706 213,196 32 General Surgery 37,604 40,605 44,473 48,805 53,641 43 Obstetrics/Gynecology 43,068 44,547 46,168 47,802 48,962 14 Otolaryngology 9,839 10,326 10,877 11,520 12,248 24 Orthopedic Surgery 23,225 24,804 26,736 28,965 31,596 36 Urology 10,690 11,455 12,448 13,696 15,122 41 Ophthalmology 18,876 20,099 21,650 23,643 25,972 38 Other Surgical Specialties 17,858 19,296 21,167 23,276 25,655 44 Other Patient Care 206,545 221,355 239,224 258,315 280,405 36 Psychiatry 44,495 46,877 49,340 51,537 54,116 22 Anesthesiology 36,762 39,547 43,188 47,499 52,493 43 Emergency Medicine 23,494 24,813 26,206 27,802 29,505 26 Radiology 30,354 33,218 36,919 41,005 45,855 51 Pathology 16,757 18,229 20,174 22,019 24,167 44 Other Specialties 54,683 58,672 63,398 68,453 74,270 36 Non Patient Care 47,940 51,182 55,093 59,422 64,394 34 Total U.S. Population (Thousands) 281,422 294,100 307,075 321,107 335,444 19

 

	Exhibit 5.4 Forecasts of Physician Requirements Under the Status Quo Scenario
	Exhibit 5.4 Forecasts of Physician Requirements Under the Status Quo Scenario (Text Only)
	2000 2005 2010 2015 2020 General Primary Care 268,710 283,632 300,651 320,992 344,907 Medical Specialties 96,926 104,145 113,200 123,560 135,331 Surgery 161,160 171,133 183,519 197,706 213,196 Other Patient Care 206,545 221,355 239,224 258,315 280,405 Non Patient Care 52,327 55,807 59,981 64,626 69,979

	Scenario 2, baseline, produces the requirements forecasts that are most likely to occur based on projected trends in managed care growth and the shifting of care from higher cost to lower cost settings. This scenario is comparable to the baseline scenario in the NDM, described later in Section 5.2, which assumes that HMO enrollment rates will increase by half a percentage point per year between 2000 and 2020 (with the gains in HMO enrollment coming from the population insured under a fee-for-service arrangement). In addition, this scenario assumes that each year, 2 of inpatient-based surgeries will shift to an outpatient setting. Regression analyses conducted to update the NDM find that for each 1 increase in the proportion of hospital-based surgeries performed on an outpatient basis, demand for inpatient days at acute care hospitals will decline by 0.47, outpatient visits will increase by 0.66, and home health visits will increase by 0.86. Using this information, the baseline scenario assumes a gradual decrease in per capita demand for inpatient days and surgery performed on an inpatient basis, and a gradual increase in outpatient visits and "other" visits. Exhibit 5.5 presents the forecasts for this scenario. Under this scenario, total requirements for physicians would increase by 28 percent between 2000 and 2020 to 996,400. Compared to the status quo scenario, there would be the same level of growth in general primary care specialties (28 percent), but slower growth in medical specialties (33 percent versus 40 percent), surgical specialties (17 percent versus 32 percent), and "other" patient care specialties (32 percent versus 36 percent).

	Exhibit 5.5 Impact of Changing Demographics on Requirements for Physicians: Baseline Scenario
	Specialty 2000 2005 2010 2015 2020 % Change 2000 to 2020 Total Physicians (MDs and DOs) 781,282 823,465 874,019 931,208 996,387 28 Total Patient Care Physicians 733,342 772,936 820,389 874,069 935,248 28 General Primary Care 268,710 284,113 301,283 321,556 345,039 28 GP & FP 109,571 115,576 122,428 130,168 138,846 27 General Internal Med. 106,411 114,438 123,929 134,583 146,730 38 Pediatrics 52,728 54,099 54,926 56,806 59,463 13 Medical Specialties 96,926 102,850 110,381 119,005 128,730 33 IM Subspecialties 40,205 42,759 46,041 49,799 53,993 34 Cardiovascular Diseases 20,828 22,235 24,192 26,629 29,440 41 Other Medical Specialties 35,893 37,856 40,149 42,577 45,297 26 Surgery 161,160 165,957 172,525 180,173 188,291 17 General Surgery 37,604 38,974 40,943 43,086 45,378 21 Obstetrics/Gynecology 43,068 43,721 44,495 45,260 45,567 6 Otolaryngology 9,839 10,003 10,214 10,498 10,847 10 Orthopedic Surgery 23,225 23,995 25,001 26,169 27,547 19 Urology 10,690 11,115 11,737 12,567 13,511 26 Ophthalmology 18,876 19,746 20,915 22,491 24,378 29 Other Surgical Specialties 17,858 18,402 19,219 20,102 21,064 18 Other Patient Care 206,545 220,016 236,199 253,334 273,187 32 Psychiatry 44,495 46,925 49,329 51,398 53,782 21 Anesthesiology 36,762 39,547 43,188 47,499 52,493 43 Emergency Medicine 23,494 24,285 25,103 26,079 27,122 15 Radiology 30,354 33,218 36,919 41,005 45,855 51 Pathology 16,757 18,229 20,174 22,019 24,167 44 Other Specialties 54,683 57,812 61,487 65,333 69,768 28 Non Patient Care 47,940 50,528 53,630 57,140 61,139 28 Total Population (Thousands) 281,422 294,100 307,075 321,107 335,444 19

	Scenario 3, universal health care coverage, assumes that the entire U.S. population has medical insurance. Under this scenario, the PARM moves a portion of the uninsured population into the insured fee-for-service and HMO settings based on the current proportion of the insured population in each of those two settings. The primary motivation for this scenario is that some advocates for the uninsured would like to see the Government sponsor more initiatives to cover the uninsured. Under this scenario, total demand for physicians would have been an estimated 817,615 in 2000, and would increase to an estimated 1,092,400 in 2020-a 40 percent increase from current (2000) baseline and/or status quo levels (Exhibits 5.6, 5.7, and 5.8). (It should be noted that under the status quo scenario, although substantially short of universal coverage, the percentage of population with medical insurance will rise over time as the population ages and a larger proportion of the population becomes Medicare-eligible.) Scenario 4 is universal health care coverage with 100 of the population enrolled in a health maintenance organization. The motivation for this scenario is work performed by Weiner (1994) and others on requirements for physicians under a managed care environment. Under this scenario, total physician requirements would have been an estimated 781,900 in 2000 and would increase to 1,059,900 in 2020-a 36 percent increase from current levels (Exhibits 5.6, 5.7, and 5.8). Scenario 5, non-minority rates, assumes that minorities have similar rates of medical insurance coverage as non-Hispanic whites within each demographic group defined by age and sex. Under this scenario the percentage of the population uninsured, insured under a fee-for-service arrangement, and in an HMO applicable to non-Hispanic whites is applied to the other two race/ethnicity groups. The motivation for this scenario is equality across racial and ethnic groups in access to medical coverage. Under this scenario, demand for physicians would have been an estimated 802,400 in 2000, increasing to 1,072,000 in 2020-a 37 percent increase from current levels (Exhibits 5.6, 5.7, and 5.8).

	Exhibit 5.6 Forecasted Physician Requirements Under Five Scenarios
	Specialty Scenario 1: Status Quo Scenario 2: Baseline Scenario 3: Universal Coverage Scenario 4: 100HMO Scenario 5: Non-minority Rates 2000 2020 2000 2020 2000 2020 2000 2020 2000 2020 Total Physicians (MDs and DOs) 781,282 1,038,234 781,282 996,387 817,615 1,092,381 781,889 1,059,907 802,356 1,072,048 Total Patient Care Physicians 733,342 973,840 733,342 935,248 767,420 1,024,551 735,131 995,794 753,007 1,005,383   General Primary Care 268,710 344,907 268,710 345,039 281,421 362,692 294,546 384,349 275,137 355,105     GP & FP 109,571 139,252 109,571 138,846 112,652 143,979 107,348 139,754 111,543 142,514     General Internal Med. 106,411 146,885 106,411 146,730 112,347 154,794 121,979 170,451 108,661 150,594     Pediatrics 52,728 58,770 52,728 59,463 56,422 63,919 65,219 74,145 54,933 61,997 Medical Specialties 96,926 135,331 96,926 128,730 101,452 141,911 101,499 144,317 99,016 138,814     IM Subspecialties 40,205 56,955 40,205 53,993 42,540 60,263 41,819 59,922 41,176 58,572     Cardiovascular Diseases 20,828 31,690 20,828 29,440 21,549 32,742 21,130 32,944 21,110 32,183     Other Medical Specialties 35,893 46,687 35,893 45,297 37,364 48,906 38,550 51,450 36,729 48,058 Surgery 161,160 213,196 161,160 188,291 170,919 226,585 161,462 218,836 165,223 219,711     General Surgery 37,604 53,641 37,604 45,378 39,463 56,421 37,187 54,102 38,479 55,185  Obstetrics/Gynecology 43,068 48,962 43,068 45,567 48,880 56,185 44,818 51,627 44,939 51,730     Otolaryngology 9,839 12,248 9,839 10,847 10,294 12,856 7,985 10,281 10,035 12,557     Orthopedic Surgery 23,225 31,596 23,225 27,547 24,741 33,604 22,066 30,884 23,803 32,534     Urology 10,690 15,122 10,690 13,511 11,175 15,755 11,236 16,260 10,824 15,341     Ophthalmology 18,876 25,972 18,876 24,378 18,704 25,915 20,194 28,701 19,034 26,254     Other Surgical Specialties 17,858 25,655 17,858 21,064 17,663 25,850 17,977 26,980 18,110 26,110 Other Patient Care 206,545 280,405 206,545 273,187 213,627 293,363 177,624 248,293 213,631 291,753     Psychiatry 44,495 54,116 44,495 53,782 42,133 52,752 31,961 38,129 45,871 56,328     Anesthesiology 36,762 52,493 36,762 52,493 38,804 55,424 38,804 55,423 37,528 53,781     Emergency Medicine 23,494 29,505 23,494 27,122 23,606 30,114 13,086 17,918 24,378 30,766     Radiology 30,354 45,855 30,354 45,855 32,865 49,649 32,865 49,649 31,543 47,822     Pathology 16,757 24,167 16,757 24,167 17,989 26,011 17,989 26,011 17,290 25,087     Other Specialties 54,683 74,270 54,683 69,768 58,230 79,413 42,918 61,163 57,021 77,970 Non Patient Care 47,940 64,394 47,940 61,139 50,195 67,830 46,759 64,113 49,349 66,665

	Exhibit 5.7 Forecasts of Physician Requirements in 2000 Under Alternative Scenarios
	Exhibit 5.7 Forecasts of Physician Requirements in 2000 Under Alternative Scenarios (Text Only)
	Total Physicians Status Quo 781,282 Baseline 781,282 Universal Coverage 817,615 Universal HMO Coverage 781,889 Non-minority Rates 802,356