The Critical Care Workforce: 
A Study of the Supply and Demand for Critical Care Physicians
Requested by: Senate Report 108-81

Executive Summary

Patients in acute care hospitals receive over 18 million days of care in intensive care units (ICU) each year, with related health care costs estimated to be almost one percent of U.S. gross domestic product.  The demand for ICU services is projected to grow rapidly during the next decade as the average acuity of hospitalized patients rises with growth in the elderly population.   The ability of critically ill patients to receive adequate care depends upon a number of factors, including the availability of highly trained health care professionals.

Organizations such as the Leapfrog Group have promoted the increased use of critical care physicians (intensivists), in ICUs as a growing body of research finds that greater use of intensivists results in improved patient outcomes.  A report by the Committee on Manpower for the Pulmonary and Critical Care Societies (COMPACCS, Angus et al., 2000) projected a growing shortfall of intensivists unless changes are made to increase the number of physicians trained in critical care.

In June 2003, in response to concerns about the widening gap between the size of the Nation's aging baby boom population and the number of pulmonary and critical care physicians, Congress asked the Health Resources and Services Administration (HRSA) to examine the adequacy of the critical care workforce.  HRSA maintains physician workforce supply and demand models developed to assess the adequacy of supply for many physician specialties.  Working with the American College of Chest Physicians, HRSA and its consultants updated the physician workforce models to include critical care physicians.

Currently, intensivists direct the care of only one third of critically ill patients.  In recent years, however, the proportion of patients receiving care under the direction of an intensivist has increased dramatically and this trend will likely continue.  An upper bound on the demand projections assumes that intensivists direct the care of approximately two thirds of patients in the ICU, while a lower bound assumes that intensivists will continue to direct the care of only a third of critically ill patients.  Our analysis supports the findings that demand for intensivists will continue to exceed available supply through the year 2020 if current supply and demand trends continue.

The intensivist supply and demand projections reported by COMPACCS are the only recent projections available for comparison.  The COMPACCS findings also suggest a growing shortage of intensivists, but the projections differ in several respects from our more recent projections.  First, the COMPACCS study uses a broader definition of intensivists to include pulmonologists who spend part of their time providing critical care services.  Second, the baseline supply projections show little growth, whereas the number of fellows in critical care started to increase in the years between the COMPACCS study and our more recent projections. Third, the COMPACCS study identified the potential growth in demand for intensivists beyond the one third of critically ill patients currently cared for by intensivists.  Their demand projections assume that current patterns of intensive service utilization and delivery will continue in future years; growth is based entirely on the growing and aging U.S. population.  If the trend towards greater utilization of intensivists in ICUs were taken into account, the COMPACCS demand projections would likely underestimate requirements for intensivists.

Despite the differences between our projections and the COMPACCS projections, the approach used by COMPACCS is methodologically sound and the data collected through their survey provides important insights on the practice behavior and retirement patterns of physicians providing critical care services.  Many of the findings from the COMPACCS study and survey provide estimates of important parameters used in the HRSA physician workforce models.   

Vulnerable populations, particularly the uninsured and those located in rural areas, likely have limited access to intensivist services.  Moreover, many critically ill Americans may receive less than the evolving standard of care because of an inability of smaller hospitals and those serving vulnerable populations to sustain ICUs in which intensivists are always available to direct care. Because a large proportion of critical care fellows are international medical graduates (IMGs), the shortage of intensivists is worsened by an inability of many qualified IMG intensivists to remain in the United States because of visa restrictions (data on the actual numbers of IMGs who return to their home countries are not available).  Furthermore, the profession has had difficulty attracting qualified applicants from U.S. medical schools and retaining practicing physicians.  We note that our study focuses on the adequacy of intensivist supply to provide adult critical care.  Population projections suggest a large increase in demand for such services due to an aging population.

Specific Findings and Conclusions:

The supply of intensivists (those physicians who identify themselves primarily as critical care physicians and have completed a critical care fellowship) will likely grow by about 48 percent—from 1,900 to 2,800—between 2000 and 2020 if current supply trends continue.  While the focus of this study is intensivists, some pulmonologists also provide critical care.  Such pulmonologists average spending less than a quarter of their time in ICUs.  HRSA's physician workforce models separately track supply and demand for pulmonologists, but do not track the amount of critical care provided by pulmonologists.  This study assumes that pulmonologists will continue to provide the same proportion of critical care estimated by the COMPACCS study.   

• Because critical care is a relatively new and growing specialty, the intensivist workforce is relatively young.  Around 2020, the intensivist workforce will likely stabilize as the number of intensivists retiring will approximately equal the number of new intensivists.
• The growth and aging of the population alone will increase demand for adult intensivist services by approximately ­38 percent—from 1,900 to 2,600 between—2000 and 2020. This increase represents a lower bound on expected growth in demand and assumes that intensivists continue to treat only a third of critically ill patients.  Further, it assumes that supply and demand were in balance in 2000.
• If the proportion of ICU patients whose care is directed by an intensivist were to increase from one third to a more optimal level of two thirds, then intensivist requirements would grow from a need for 3,100 FTE intensivists in 2000 to 4,300 by 2020.  This  represents  a shortage of about 1,200 intensivists in 2000, growing to 1,500 in 2020, or 129 percent above the projected supply.  The upper bound on the demand projections reflects the large potential growth in utilization of intensivist services—especially in metropolitan areas.
• Lifestyle issues associated with critical care as it is currently practiced present a barrier to increasing the number of practicing intensivists.  Reimbursement for critical care is also perceived by those in the profession as inadequate, making critical care less attractive to newly trained physicians.
• Critical care remains an evolving specialty.  A significant body of literature indicates that the current supply of practicing intensivists is lower than what is required to care for patients in U.S. ICUs.  The evidence indicates that patient outcomes are improved when intensivists are available around-the-clock for patient consultation.
• Organizational changes in the way that care is provided to critically ill patients have the potential to improve patient access to cost-effective and quality care—especially in rural areas.  One example is the increased use of electronic ICUs where specialist physicians and nurses monitor and help treat critically ill patients in widely scattered hospitals.

In summary, we project that if current trends continue, the growing supply of intensivists will be insufficient to provide the optimal level of care to future populations through 2020.  A lower bound of projected demand assumes that all growth in demand for intensivist services is due to the growth and aging of the population but the recent growth in intensivist involvement in ICU care suggests that this lower estimate is highly unlikely.  Total employment opportunities will likely grow faster than this lower bound as hospitals increasingly staff their ICUs with intensivists.  An upper bound on the demand projections would occur if intensivists direct the care of two thirds of patients admitted to the ICU.  The likely demand for intensivists will likely lie somewhere between this upper and lower bound, suggesting the need to increase intensivist supply and to continue monitoring trends in supply and demand.  

Introduction

The sickest patients in U.S. hospitals are cared for in intensive care units (ICU).  The number of patients cared for in ICUs is projected to grow rapidly during the next decade as the average acuity of hospitalized patients rises with growth in the elderly population, who consume the greatest amount of health care services.  In the U.S., patients in ICUs currently receive over 18 million days of care every year, with related health care costs estimated to be almost one percent of U.S. gross domestic product. ^[1]  The ability of these critically ill patients to receive appropriate care depends upon access to hospitals with appropriate facilities as well as the availability of highly trained health care professionals.

In 2000, Derek Angus and his colleagues on the Committee on Manpower for the Pulmonary and Critical Care Societies (COMPACCS) presented a detailed analysis of the critical care physician (or “intensivist”) workforce that provides care to patients in ICUs. ^[2]  The report, published in the Journal of the American Medical Association (JAMA), reviewed the data supporting the increasing likelihood of a shortage of physicians adequately trained to care for the sickest patients in the U.S. health care system.  The COMPACCS study concluded that, if current trends in the utilization and supply of intensivist services continued, a severe shortage of intensivists would materialize within the next decade.  The study also found that two-thirds of critically ill patients did not receive care from intensivists (which in the COMPACCS study also includes pulmonologists who spend some time in ICUs). 

Medicine and medical care has evolved considerably over the last century in both inpatient and outpatient settings.  Primary care physicians generally directed care for patients inside and outside of hospitals in consultation with appropriate specialists.  However, the greater trend towards specialization of the physician workforce over time has changed the roles of both generalist and specialist physicians.  In the inpatient setting, the physician directing care (“attending”) during a course of hospitalization has traditionally guided this care regardless of the unit in which a patient is cared for, including the ICU.  A 1998 survey of U.S. ICUs found that the care of critically ill patients was directed by full-time intensivists for 23 percent of patients and intensivists were consulted for another 14 percent; other patients were cared for by primary care physicians who have not completed critical care fellowships. ^[3] Though it is unclear why intensivists do not direct care for more critically ill patients, Buchardi and Moerer suggest that this may be due to the fact that “many primary physicians resist relinquishing authority for their patients, and intensivists may tend to exclude the primary physicians from decision-making.” ^[4]

The COMPACCS study found that the demand for critical care services would increase rapidly due to the aging and expanding of the population.  Meanwhile, the supply of physicians trained to provide these services would remain constant through 2030, making it unlikely that intensivists would be able to care for a greater proportion of critically ill patients.  In addition, the growing body of literature linking full-time intensivist staffing with improved outcomes for ICU patients has increased the demand for physicians trained in critical care.  Hospitals have also been encouraged to make organizational changes to their ICUs by employer groups and other payers of health care.    

In June 2003, in response to concerns about the widening gap between the size of the Nation's aging baby boom population and the number of pulmonary and critical care physicians, Congress asked the Health Resources and Services Administration (HRSA) to examine the adequacy of the critical care workforce.  HRSA maintains several health workforce models to assess the adequacy of future physician and nurse supply in different specialties and settings.  Working with the American College of Chest Physicians (ACCP), HRSA updated its physician workforce models to add critical care as a separate specialty.  The purpose of this report is to assess the current and future adequacy of supply of critical care physicians.  Our analysis supports the findings that demand for intensivists will continue to exceed available supply through the year 2020 if current supply and demand trends continue.

The Critical Care Physician Workforce and Physician Modeling

According to the American Medical Association (AMA), over 750,000 allopathic and osteopathic physicians were actively practicing medicine in the U.S in 2003, with fewer than 5,000 trained and certified in critical care.  The two major research questions guiding this study are (1) do we currently have a sufficient supply of intensivists, and (2) will supply be sufficient over the next decade or two?

The factors affecting the supply of, and demand for, physician services are complex and dynamic.  In this report we provide a brief description of the assumptions, methods and data used to project the future supply of and demand for intensivist services.  A more detailed description of HRSA’s workforce models is available in other reports. ^[5]

Workforce projections provide an indication of the magnitude of likely imbalances in supply and demand in future years, and thus are useful for planning purposes.  Projections of physician shortages and surpluses have influenced policies and programs for over 100 years, helping to determine the number and specialty composition of physicians being trained. ^[6] The Flexner Report in the early 1900's is considered the first major attempt to systematically analyze the adequacy of the physician workforce; the one outcome of this study was a relative downsizing of the physician workforce between 1900 and 1930, with a decrease from 175 to 125 physicians for every 100,000 persons. ^[7] In 1932, a national Commission on Medical Education called for a further reduction in the size of the physician workforce.  In the late 1950’s, the Bayne-Jones and Bane reports from the U.S. Office of the Surgeon General projected an impending physician shortage.  By 1960, immigration restrictions on physicians were relaxed.  This was followed by an expansion of the Nation’s medical schools, an increase in government funding for medical education, and the creation of policies and programs that encouraged immigration of foreign-trained physicians.  Efforts to increase the physician supply were so successful that by the late 1970's experts were predicting a growing oversupply of physicians.

Though the Graduate Medical Education National Advisory Committee in 1980 projected a surplus of physicians (particularly specialists) by 2000, the number of physician trainees continued to rise until the early 1990’s.  The expanding enrollment in health maintenance organizations (HMOs) during the 1980's and 1990's prompted re-examination of the adequacy and composition of the physician supply.  Subsequent models and their projections assumed that the U.S. would move quickly to a more primary care-oriented system with more efficient delivery of health care services.  As a result, most models predicted that the United States would have a large surplus of specialists by 2000. ^[8] New medical graduates became less likely to enter fellowships after residency training, particularly those in internal medicine. ^[9] By the late 1990’s, however, the trend towards greater specialization returned.  In 1998, only 43 percent of residents in internal medicine went on to subspecialize; by 2003, that proportion had increased to over 66 percent. ^[10]

In contrast to the widely held consensus of the mid-1990's that the United States would have a surplus of specialists, a growing number of researchers have suggested that growth in the number of specialists, ^{[11], [12], [13]} not primary care physicians, will be especially important in meeting the demands of an aging and expanding population. ^{[14], [15]}  The recent discussions regarding the adequacy of the future supply of physicians have centered around the theory that economic growth is a major determinant of growth in per capita demand for physician services, and that continued economic expansion will contribute to a significant shortage of physicians—particularly specialists—over the next decade.  While these theories are still being debated in the literature, new concerns about the shortage of physician specialists have rarely informed the debate about the demand for individual specialties with detailed analysis or projections.

Few analyses have rigorously examined the adequacy of physician supply in critical care.  In the late 1990’s, the American Thoracic Society (ATS), the American College of Chest Physicians (ACCP), and the Society of Critical Care Medicine (SCCM) formed the Committee on Manpower for the Pulmonary and Critical Care Societies which examined the supply of, and demand for, intensivists and pulmonologists.  The study projected a large increase in demand after 2007 (based primarily on the aging of the U.S. population) and relative shortages in the supply of these physician specialists.

COMPACCS estimated a shortage in the number of available intensivist hours of care equal to 22 percent of demand by 2020 and 35 percent by 2030. ^[16] In their analysis, the shortage became more severe if the demand for intensivist care was extended to a greater proportion of ICU patients.  Alternative scenarios modeling changes in the variables affecting demand for critical care services, including greater managed care penetration, had little impact on this shortage. 

This shortage was projected based upon best available data at the time of the study, including the number of physicians choosing pulmonary and critical care specialties, and the expectation that these numbers would remain stable over time.  The study also tested several scenarios affecting both supply (changing assumptions about number of hours worked, age of retirement, distribution of time between clinical and other activities) and demand (changing assumptions about penetration of managed care, growth in outpatient care, and other factors).  The COMPACCS study anticipated many of the changes in the workforce and the delivery of health care, though some changes (such as the prominent role of intensivists and critical care in the quality movement) were unforeseen.

The methods and assumptions used in the COMPACCS study are similar to those used in the Physician Supply Model (PSM) and Physician Demand Model (PDM) developed by HRSA. Historically, the PSM and PDM modeled the supply of, and demand for, pulmonologists separately from other specialties, but grouped intensivists with several other smaller specialties into its “other internal medicine specialties” category.  The revised PSM and PDM have expanded the number of individual physician specialties modeled, creating the capability to project the future supply of and demand for intensivists.

Using findings from the literature, original research, and projections from the PSM and PDM, this report examines the current and future supply of physicians who provide care to critically ill patients; the major factors and trends affecting the demand for their services; and the likely inadequacy of their numbers through 2020.  Chapter 1reviews the history of intensivist training and practice, the issues related to the critical care workforce, and trends in the organization and delivery of critical care.  Chapter 2 discusses the supply of critical care physicians, the factors affecting the availability of practicing specialists, and the projected supply of critical care physicians through the year 2020.  Chapter 3analyzes the demand for critical care physicians, its determinants, and the ratio of intensivists to population required to meet the demand for related services.  The chapter concludes with projections of demand for critical care physicians through 2020.  Chapter 4 compares the current projections with those of the COMPACCS study and discusses the implications of these results.  Chapter 5 examines the implications for vulnerable populations, provides examples of how unmet demand has been addressed, and suggests areas for future research.

ACKNOWLEDGEMENTS

This publication was prepared for the Health Resources and Services Administration by the Lewin Group under Contract  Number  HRSA-250-01-0001.  Staff from HRSA's National Center for Health Workforce Analysis contributed to writing the report.