The Critical Care Workforce: A Study of the Supply and
Demand for Critical Care Physicians
Chapter
2: Supply
Historically,
there has been greater consensus on physician supply projections
than for demand projections. Projecting the future supply
of active physicians is relatively straightforward, and
is accomplished by adding annual estimates of newly trained
physicians to current supply, and subtracting estimates
of the number of physicians retiring. While physician
supply refers to the number of active physicians, effective
supply refers to the amount of services provided expressed
as full-time equivalent (FTE) physicians. Projecting effective
physician supply is more challenging due to the incomplete
information on physician activity and behavior. The personal
choices made by physicians determine the number of hours
spent providing care, medical specialties chosen, productivity,
work location, and retirement behavior.
Physician
productivity is influenced by more than physician preferences,
but also depends on external factors such as the activity
of other health workers. Changes in the use of non-physician
clinicians and other health professionals, technological
advances, epidemiologic trends, amount of time spent with
patients per visit, and changes in the health care operating
environment all affect both the average number and type
of patients seen per physician. For instance, the average
number of patient visits declined during the 1990's, due
mainly to a decline in inpatient activity, with office visits
per physician remaining relatively constant.
The
Current Supply of Physicians Trained in Critical Care
In 2000,
the base year for this analysis, 10,360 physicians reported
their primary medical specialty as critical care or pulmonology
as recorded in the AMA Masterfile. About 65 percent of
those physicians report pulmonary medicine as their primary
specialty. Among those whose practice includes critical
care, 19 percent are dual trained in critical care and pulmonology
(CCP), 10 percent are internists trained in critical care
medicine (CCM) alone, and the remaining 6 percent are divided
evenly between critical care anesthesiologists (CCA) and
critical care surgeons (CCS). [34]
For the purposes of this study, physicians are considered
“intensivists” when they have received primary training
in medicine, surgery, or anesthesiology, as well as 2-3
years of training in critical care medicine. [35]
Exhibits 1 and 2show the number of active physicians, by
specialty, between 1998 and 2001. [36]
As is
evident from Exhibit 2, the number of physicians self-designated
as practicing “pulmonary/critical care medicine” nearly
doubled in the 3-year period from 1998 to 2001. [37]
This may reflect a shift in training programs from pulmonary
medicine alone to combined pulmonary and critical care programs
as well as a change in self-designation choices. However,
self-designated specialty does not reveal how physicians
are actually spending their clinical time. That is, whether
a physician has completed a program in “pulmonary/critical
care medicine” does not guarantee that he or she will spend
any clinical time practicing as an intensivist.
Exhibit
1. Number of Intensivists by Primary, Self-Designated Specialty [D]
Exhibit
2. Number of Self-Designated Pulmonologists and Critical
Care Pulmonologists
 [D]
Clinical
Activity
On average,
94 percent of pulmonologists and critical care physicians
were engaged primarily in direct patient care (as opposed
to administrative work, research, teaching, or other work)
in 2000. CCP physicians were the most likely to be engaged
in patient care (98 percent), versus 88 percent of those
who considered themselves pulmonary specialists alone.
Respondents to the 1997 COMPACCS survey worked an average
of 61 hours per week and spent about a quarter (26 percent)
of their time in the ICU. This proportion was less for
those with pulmonary training (23 percent) and twice as
high for those with training in critical care only (46 percent).
Gender
While
women are gaining representation in critical care and pulmonary
specialties, males still comprise 86 percent of pulmonologists
and critical care physicians. The highest proportion of
females were CCPs (23 percent) and CCAs (19 percent). Men
comprised a higher proportion of CCPs (90 percent) and CCMs
(83 percent). These differences may be related to the combined
trends of increased female participation in medicine and
the change in pulmonary training programs to include critical
care.
During
the past three decades, the proportion of graduates from
U.S. medical schools who are female has risen from 10 percent
to about 50 percent. Because work and retirement patterns
differ systematically by gender, the increasing proportion
of physicians who are female has profound implications for
the overall supply of physician services. Female physicians
tend to work approximately 15 percent less time in patient
care then do their male counterparts after controlling for
age and specialty. Female physicians are more likely than
their male counterparts to choose non-surgical specialties,
spend fewer hours providing patient care, are less likely
to work in rural areas, and tend to retire earlier. The
COMPACCS data indicate that female physicians practicing
as intensivists or pulmonologists worked an average of 300
hours less per year than their male counterparts.
Age
The
majority of critical care and pulmonary physicians are between
35 and 44 years of age, reflecting the relatively new status
of both specialties. Self-designated pulmonologists tend
to be older than physicians practicing exclusively critical
care. Approximately 64 percent of pulmonologists are over
the age of 45 as compared to between 4 percent and 31 percent
of the physicians for each of the other critical care specialties
examined. An older cohort of pulmonologists may be accompanied
by a greater likelihood of their retirement in the near
future as compared to other critical care physicians. Female
physicians were, on average, 2 years younger than their
male counterparts, reflecting the growing trend toward feminization
of the medical workforce. Approximately 46 percent of male
physicians were 45 years or older versus 23 percent of females
(Exhibits 3 and 4). Age is significant because it is highly
correlated with retirement decisions and plays a significant
role in hours worked. Physicians over the age of 65 tend
to work fewer hours than younger physicians.
Exhibits
3 and 4. Age Distribution of Physicians in Pulmonary and
Critical Care Specialties, by Gender
 [D]
Source:
COMPACCS data.
 [D]
Source:
COMPACCS data.
New Entrants
to the Critical Care Workforce
Physicians
in the United States enter the workforce after completing
the requirements for licensure in individual States. These
requirements differ by location, but include the completion
of a medical degree (a Doctorate of Medicine [MD] or Doctorate
of Osteopathy [DO]) as well as the completion of post-graduate
medical education (GME) training in an internship and residency
program that ranges from 1 to 8 years. Schools of allopathic
medicine graduate approximately 16,000 MDs each year. This
number has been relatively stable since 1980. Schools of
osteopathic medicine graduated approximately 2,600 DOs in
2001 and this number has been steadily increasing in recent
years.
Almost
30,000 physicians completed their GME training and became
eligible to practice a chosen specialty in 2004. [38]
Physicians in non-surgical subspecialties (i.e., outside
of family practice, general internal medicine, and general
pediatrics) must complete an initial residency period before
entering subspecialty training (fellowships). In 2004,
22,444 physicians were scheduled to start GME programs for
the first time, the highest number on record.
Almost
one-fourth of physicians in GME training programs are International
Medical Graduates (IMG) who received their medical degrees
abroad. Many of the 5,000 IMGs who enter U.S. GME programs
each year do so under the temporary work (H) or training
(J) visa programs. IMGs may remain in the United States
after completing training if they are citizens or permanent
residents (U.S. IMG) who graduated from medical schools
in other countries. In addition, foreign IMGs can participate
in the J-1 Visa Waiver Program which waives the requirements
that foreign physicians return to their country for a minimum
of 2 years before practicing in the U.S. This waiver is
granted in exchange for a commitment to deliver primary
care services to underserved communities.
The
training of physicians in critical care medicine may take
10 or more years of graduate training, including 4 years
of medical school, 3 or more years of residency, and 2 or
more years of fellowship training in critical care (medicine,
anesthesia, or surgery) or pulmonary/critical care. In
2003, 86 physicians completed training in critical care
(internal medicine), 57 completed pulmonary (internal medicine),
and 359 completed combined pulmonary-critical care programs.
[39] In 1996,
the COMPACCS group reported 354 graduates from pulmonary
and pulmonary/critical care medicine training programs;
110 from critical care internal medicine programs; and 130
graduates from critical care programs in departments of
anesthesiology and surgery (63 and 67, respectively). In
the year 2002, there were 1,374 fellows in all critical
care and pulmonary training programs. A majority (72 percent)
of those residents were in combined pulmonary/critical care.
Even if all physicians with some critical care training
were to deliver critical care services, less than one percent
of U.S. medical school graduates are expected to choose
to practice as intensivists. Moreover, the number of filled
fellowship positions in CCA, CCM, and CCS has fallen since
1995 (Exhibits 5 & 6). [40]
In CCM alone, the number of current fellows has dropped
by over 25 percent. While the number of physicians graduating
has grown slightly over time, the number of newly trained
critical care medicine fellows has dropped from 110 (1998)
to 86 (2004) per year.
Exhibit
5. Filled Fellowship Slots in Critical Care
 [D]
Exhibit
6. Trainees in Pulmonary and Pulmonary/Critical Care Fellowships
 [D]
Other
factors may also affect the future effective supply of intensivists,
including the proportion of IMGs who fill fellowship positions.
In 2003, IMGs accounted for 67.4 percent of fellows in critical
care (anesthesia); 18.9 percent of critical care (surgery)
fellows; and 67.8 percent of those in critical care (internal
medicine) (Exhibit 7). By 2003, 79.4 percent of pulmonary
(internal medicine) fellows and 38.1 percent of those in
pulmonary/critical care programs were also IMGs (Exhibit
8). The country of medical school training is important
because, although almost half of IMGs are actually U.S.
citizens or permanent residents, physicians who train on
J‑1 or other visitor visas may be required to return
to their country of citizenship unless they are granted
a visa waiver. Relatively few physicians practicing outside
of primary care qualify for such waivers.
Fewer
residents are entering pulmonary fellowships alone (without
critical care) with more receiving at least some part of
their training in critical care. For this reason, it is
more revealing to examine the number of new board certifications
in critical care as displayed in Exhibit 9. The
number of new board certified critical care specialists
declined by almost half, from 1,135 to 660 new diplomats.
This number excludes physicians trained in pulmonary medicine
alone.
Exhibit
7. Percent of Critical Care Fellows who are IMGs
 [D]
Exhibit
8. Percent of Fellows in Pulmonary and Combined Pulmonary—Critical
Care Training Programs who are IMGs
 [D]
Exhibit
9. New Pulmonary and Critical Care Certifications, 1991-2001
 [D]
Retirement
of Critical Care Physicians
Physicians
leave the workforce through retirement, mortality, disability,
and career change. An accurate estimate of separation rates
is crucial for projecting physician supply. Historically,
estimates of physician retirement rates have come from analysis
of the AMA Masterfile data.
The
high stress of working in the ICU may contribute to earlier
retirement by intensivists. A study measuring the prevalence
of burnout in critical care examined the levels of exhaustion
in a sample of members from the internal medicine section
of the Society for Critical Care Medicine, over half of
whom worked more than 50 percent of their time on critical
care. [41] The
authors report that a third of the respondents scored in
the high range for emotional exhaustion and a fifth scored
in the high range for depersonalization. In addition, over
half scored in the low range for personal achievement.
Original COMPACCS survey data reflects the tendency of intensivists
to retire at earlier ages than pulmonologists (Exhibit 10).
Over one half of intensivists expect to retire by the age
of 60 and almost a third expects to retire by the age of
55.
Exhibit
10. Retirement Expectations of Pulmonary & Critical
Care Physicians
 [D]
Physician
Supply Model Projections
All
of the factors described above impact the “effective” supply
of physicians practicing as intensivists. Part-time intensivist
practice, whether associated with age, gender, or primary
specialty training area, effectively reduces the number
of full-time equivalent (FTE) physicians available. For
example, adding 2,000 physicians that practice as intensivists
50 percent of the time to a base of 2,000 full-time intensivists
would deliver the amount of services associated with 3,000
(not 4,000) full-time intensivists.
Exhibit
11 incorporates the various elements of supply described
above. Current projections of intensivist supply indicate
that if current supply patterns continue, the effective
supply will likely increase by approximately 48 percent
between 2000 and 2020, from approximately 1,880 to 2,770,
at which time the supply becomes stable. Projections beyond
2020 are unlikely to be useful given their uncertainty.
Within the next 20 years, it is also likely that a plurality
of current intensivists will retire as a large portion of
the current supply is now between the ages of 35 and 44.
Despite an overall increase in the number of graduates with
critical care training in recent years, decreasing hours
worked and steadily rising numbers of retirements will lead
to an essentially flat number of critical care providers
by 2020.
Exhibit
11. Projected “Effective” Supply of Adult Intensivists
 [D]
|
