Original Research
Physical Therapy Workforce in the United States:
Forecasting Nationwide Shortages
Janice L. Zimbelman, PhD, MS, PT, Stephen P. Juraschek, BA,
Xiaoming Zhang, PhD, Vernon W.-H. Lin, MD, PhD
Objective: To examine current and future physical therapy (PT) job surplus/shortage
trends across the United States.
Design: Forecast models and grading methodology previously published for nursing were
used to evaluate individual state PT job shortages from 2008 to 2030.
Setting: Not applicable.
Participants: Not applicable.
Methods: The forecast model used to project PT job supply and demand accounted for
changes in age and population size on the basis of estimates from the U.S. Census Bureau for
each of the 50 states. PT shortages were assigned letter grades on the basis of shortage ratios
(difference between demand and supply per 10,000 people) to evaluate PT shortages and
describe the changing PT workforce in each state.
Results: On the basis of current trends, demand for PT services will outpace the supply of
PTs within the United States. Shortages are expected to increase for all 50 states through
2030. By 2030, the number of states receiving below-average grades for their PT shortages
will increase from 12 to 48. States in the Northeast are projected to have the smallest
shortages, whereas states in the south and west are projected to have the largest shortages.
Conclusion: These data serve to provide health professionals, policy makers, and stakeholders with a means of assessing current and future PT needs. Discussion of the issues
surrounding PT shortages and ongoing assessment of supply and demand must ensue to
mitigate projected shortages. Although our model has several limitations and may be
oversimplified, it is the first attempt to use available, creditable data to examine both supply
and demand for the entire country. Follow-up studies that use more complex modeling are
needed to adequately forecast future trends beyond that accomplished in the current article.
Monitoring trends over time is critical to maintain an appropriate balance between PT
supply and demand that meets the population needs.
PM R 2010;2:1021-1029
INTRODUCTION
Recently, US News and World Report rated physical therapy (PT) as one of the 30 best careers,
reporting that, next to clergy, PT is ranked highest in job satisfaction [1], making it a
desirable profession. The Bureau of Labor Statistics (BLS) reported that the employment of
physical therapists (PTs) is expected to grow 27% from 2006 to 2016, faster than the average
for all occupations [2]. This growth is attributed to a number of factors, including the
rapidly increasing elderly population, advances in medical technology permitting treatment
of an increasing number of previously untreatable disabling conditions, new treatments and
techniques expanding the scope of PT practice, and widespread interest in health promotion
[2]. This projected employment growth suggests that the demand for PT services will
increase; but will the supply of PTs keep pace with projected demands?
Human resource planning involves projecting the future supply of resources, in this case
PTs, and the future requirements or demands for that service. The major objectives of this
planning are to measure PT supply and PT demand, assess the balance between PT supply
and PT demand, and project the likelihood of an appropriate balance in PT supply and PT
demand [3-5]. There is a dearth of information regarding human resources in PT. The 1995
PM&R
1934-1482/10/$36.00
Printed in U.S.A.
J.L.Z. Cognitive and Motor Learning Laboratory, Research Service Louis Stokes Department
of Veterans Affairs Cleveland Medical Center; and
Healthcare First Study Group, Department of
Physical Medicine and Rehabilitation, Neurological Institute, Cleveland Clinic, Cleveland, OH
Disclosure: nothing to disclose
S.P.J. Physical Medicine and Rehabilitation,
Louis Stokes Department of Veterans Affairs
Cleveland Medical Center, Cleveland, OH;
Healthcare First Study Group, Department of
Physical Medicine and Rehabilitation, Neurological Institute, Cleveland Clinic, Cleveland,
OH; and Johns Hopkins University School of
Medicine, Baltimore, MD
Disclosure: nothing to disclose
X.Z. Physical Medicine and Rehabilitation,
Louis Stokes Department of Veterans Affairs
Cleveland Medical Center; and Healthcare First
Study Group, Department of Physical Medicine
and Rehabilitation, Neurological Institute, Cleveland Clinic, Cleveland, OH
Disclosure: nothing to disclose
V.W.-H.L. Healthcare First Study Group, Department of Physical Medicine and Rehabilitation, Neurological Institute, Cleveland Clinic,
9500 Euclid Ave. C-14, Cleveland, OH
44195; and Physical Medicine and Rehabilitation, Louis Stokes Department of Veterans
Affairs Cleveland Medical Center, Cleveland,
OH. Address correspondence to: V.W.-H.L.;
e-mail: [email protected]
Disclosure: nothing to disclose
Disclosure Key can be found on the Table of
Contents and at www.pmrjournal.org
This study was supported in part by a grant
from the California Institute for Nursing &
Health Care, and the U.S. Department of
Veterans Affairs Rehabilitation Research and
Development Service Merit Review Award.
Portions of this work were presented at the
2009 American Academy of Physical Medicine and Rehabilitation Annual Meeting.
Submitted for publication January 22, 2010;
accepted June 30, 2010.
© 2010 by the American Academy of Physical Medicine and Rehabilitation
Vol. 2, 1021-1029, November 2010
DOI: 10.1016/j.pmrj.2010.06.015
1021
1022
Zimbelman et al
volume of Physical Therapy explored issues surrounding PT
supply and PT demand in a series of articles [3,6-8]. In 1997,
the American Physical Therapy Association (APTA) commissioned a private health care study and analysis consulting
firm (Vector Research Inc) to examine the relationship between supply and demand of PTs [9]. The results of the
Vector study projected a 30% surplus of PTs by the year 2005
and continuing to 2020. However, the APTA Employment
Survey conducted in 2005 revealed that the unemployment
rate for PTs was 0.2% [8], suggesting that the predicted
surplus did not occur. Although reasons why the surplus did
not occur are not clear, presumably dramatic changes in the
market (eg, expansion of PT practice areas into nontraditional settings, PT clinicians moving into academic, administrative, or other related fields) and health care policy (eg,
direct access to PT services without physician referral, congressional moratoriums delaying implementation of the
$1500 cap on Medicare benefits imposed by the Balanced
Budget Act [BBA]) prevented the projected surplus.
Since the Vector study, only 2 other studies have examined human resource information related to PT throughout
the United States. Chevan and Chevan [10] retrospectively
examined PT supply by using publicly available census data
to develop a statistical profile of PTs. They found that the
number of employed PTs in 1990 had more than doubled
compared with the number employed in 1980. In addition,
the PT supply ratio (the number of PT jobs per 10,000
persons) had increased from 1.4 to 2.7 in the same time
period. It was not known whether this PT supply ratio of 2.7
was adequate to serve the PT needs of the population; thus,
no determination of over- or undersupply could be made.
Recently, Landry et al [11] retrospectively examined PT
supply ratios across the United States and found the national
ratio of PTs per 10,000 population increased from 3.8 in
1995 to 6.2 in 2005. It is apparent from these studies that the
PT supply ratios increased during the time periods examined.
The APTA recently renewed its efforts to examine human
resource issues within the PT profession as the result of
anecdotal reports of PT shortages. To begin this process, they
examined vacancy and turnover rates in acute care hospitals,
skilled nursing facilities, and outpatient PT offices. The results were released in a series of reports in 2008 demonstrating vacancies across the 3 health care settings [12-14]. The
vacancy rates for skilled nursing facilities, acute care hospitals, and outpatient PT offices were 18.6%, 13.8%, and
13.1%, respectively. Turnover rates of full-time equivalent
PTs were greatest for skilled nursing facilities (85.2%) [14]
and lowest for outpatient offices (8.3%) [12]. Turnover rates
for acute care hospitals were 15.9% [13]. In addition, approximately one-third of skilled nursing facilities and outpatient clinics took an average 2 to 5 years to fill a vacant
position [12,14]. These results suggest there currently is a
shortage of PTs across these 3 health care settings. Except for
the Vector study, to our knowledge, there are no other
PHYSICAL THERAPY WORKFORCE IN THE U.S.
studies examining both PT supply and PT demand for the
entire nation. In addition, we have not found published
studies that forecast future PT supply and PT demand for the
United States.
In 2005 the Healthcare First study group was established
at the University of California at Irvine to address the shortage of Registered Nurses (RNs) in the state of California. By
the use of accessible public databases, forecasting models
were constructed to project the demand and supply of RN
jobs. The combination of these supply-and-demand models
was used to produce regional RN shortage forecasts for future
years [15,16]. In 2008 the Healthcare First study group was
relocated to the Cleveland Clinic and began to address the
distribution and health care workforce at the national level by
the use of similar demand-and-supply models developed
previously. The authors used the grading methodology as
well as supply-and-demand models to examine current surplus/shortage trends in PT across the 50 states of the United
States and make predictions of these trends to the year 2030.
The results of this current study serve to provide points of
discussion for the PT profession, policy makers, and stakeholders when evaluating current and future PT needs to
mitigate predicted shortages.
MATERIALS AND METHODS
In this article, we project PT job shortages by examining the
disparity between PT demand and PT supply in all states
within the United States (see Table 1 for key term definitions). To accomplish this, forecast models were created for
both PT demand and PT supply. Estimates used by this
model came from government-maintained, readily accessible, public databases. PT job estimates were largely obtained
from the BLS, which uses a survey of employers in all industry sectors to estimate the national and state number of PT
jobs. PT jobs include both part-time and full-time workers
who are paid a salary and whose job description conforms to
the Standard Occupational Classification system used by all
federal statistical agencies to classify workers. The BLS estimates do not include self-employed workers, owners or
partners in unincorporated firms, or other unpaid workers
[17,18]. State and national population estimates and projections were obtained from the U.S. Census Bureau (USCB),
which estimates nation and state populations by incorporating birth, death, and migration rates into their decennial
survey data [19]. A grading methodology developed previously [15,16] was used as a method for evaluating individual
state shortage ratios between 2008 and 2030. States were
aggregated into 4 large regions as defined by the BLS [20] to
facilitate regional analysis (Table 2).
Demand Model
The demand model was designed on the premise that population growth and age would be the principal drivers of
PM&R
Vol. 2, Iss. 11, 2010
1023
Table 1. Explanation of key terms
Key Term
Definition
Bureau of Labor
Statistics (BLS)
The official source of labor economic and statistical data for the Federal Government. Through a
semiannual survey, the BLS produces employment and wage estimates for 800 different occupations on
the national, state, and subregional level (BLS, 2007).
Report card
A collection of grades assigned to each state on the basis of a grading rubric used for determining
stated (2008) or projected (2020 and 2030) PT shortage ratios.
National mean
5.5 PT jobs per 10,000 people. This value was determined by the number of PTs in the United States per
10,000 people for 2008 (BLS, 2008).
Personal health care An estimate that takes into account “spending for hospital care, physician and clinical services, dental
expenditure (PHE)
care, other professional services, home health care, nursing home care, and health care products
purchased in retail outlets.” This estimate does not include spending on public health programs, health
facility administration, health care research, and the construction of health care facilities (Centers for
Medicare & Medicaid Services, 2008).
Physical therapist (PT) PT is an occupation profile defined by the BLS. PTs are referred to health care professionals who diagnose
and treat individuals with medical problems that limit their ability to move and perform functional
activities. PTs examine each individual and develop a plan by the use of treatment techniques to
promote the ability to move, reduce pain, restore function, and prevent disability. In addition, PTs work
with individuals to prevent the loss of mobility before it occurs by developing fitness and wellnessoriented programs for healthier and more active lifestyles.
Physical therapy (PT) A worker who can be classified as a full-time or part-time PT. This is the fundamental unit of measure used
jobs
to estimate PT populations and is counted through a survey conducted by the BLS every 3 years.
PT demand
The estimated number of PT jobs needed to meet population needs.
PT demand ratio
The number of PT jobs needed per 10,000 people.
PT shortage
The difference between a region’s demand for PT jobs and that region’s supply of PT jobs.
PT shortage ratio
PT shortage per 10,000 people.
PT supply
The estimated number of PT jobs.
PT supply ratio
The number of PT jobs per 10,000 people.
demand for PTs in coming years. The Centers for Medicare
and Medicaid published a means of converting demographic
changes into a personal health care expenditure (PHE) [21].
By using the USCB age-population projections [22], we were
able to forecast future demand for health care services until
2030 as a single dollar amount. By means of a linear regression analysis, this monetary figure was plotted against available BLS-reported PT jobs in the United States for 2000 and
2004 to 2008, resulting in a linear slope of 3.58 ⫻ 10⫺7 and
Table 2. US regional groupings from the Bureau of Labor
Statistics [17]
South
West
Midwest
Northeast
Alabama
Arkansas
Delaware
Florida
Georgia
Kentucky
Louisiana
Maryland
Mississippi
North Carolina
Oklahoma
South Carolina
Tennessee
Texas
Virginia
West Virginia
Alaska
Arizona
California
Colorado
Hawaii
Idaho
Montana
Nevada
New
Mexico
Oregon
Utah
Washington
Wyoming
Illinois
Indiana
Iowa
Kansas
Michigan
Minnesota
Missouri
Nebraska
North Dakota
Ohio
South Dakota
Wisconsin
Connecticut
Maine
Massachusetts
New
Hampshire
New Jersey
New York
Pennsylvania
Rhode Island
Vermont
standard error of 7.91 ⫻ 10⫺9 (R2, 0.998, 95% confidence
interval, 3.36 ⫻ 10⫺7, 3.80 ⫻ 10⫺7; P ⬍ .001). This linear
slope was multiplied by the projected yearly change in PHE
of each state to produce an estimated change in PT demand
for that year. The slope standard error was used to generate
upper and lower estimates of PT demand through 2030.
With the use of 2008 values as the current PT demand, which
is an underestimate of current PT shortages according to
prior studies [12-14], the yearly projected change in PT
demand in each state was used to produce a demand estimate
up through the year 2030. The following equation describes
details of the demand model:
DR, 2009 ⫽ 5.50 ⫻ [2008 Projected State Population]/104]
⫹ 3.58 ⫻ 10⫺7 ⫻ ⌬PHER, 2008, 2009
DR,2010 ⫽ 5.50 ⫻ [2008 Projected State Population]⁄104]
⫹ 3.58 ⫻ 10⫺7 ⫻ (⌬PHER, 2008, 2009 ⫹ ⌬PHER, 2009, 2010
DR,N ⫽ 5.50 ⫻ [2008 Projected State Population]⁄104]
⫹ 3.58 ⫻ 10⫺7 ⫻
兺
R
(⌬PHE2008, 2009 ⫹ ⌬PHE2009 2010
⫹ . . . ⫹ ⌬PHEN-1,N)
where DR,N, D ⫽ demand, R ⫽ region or state, and N ⫽ year;
⌬PHEN-1,N ⫽ the change in PHE from 1 year to the next year;
and 5.50 is the national mean obtained from BLS (PT jobs per
10,000).
1024
Zimbelman et al
PHYSICAL THERAPY WORKFORCE IN THE U.S.
Supply Model
Table 3. Grading rubric used to grade projected PT shortages
ratio
Like the demand model, the supply model assumed that
changes in PT supply were determined primarily by age and
population growth. With data from the Current Population
Survey [23], which is a survey conducted jointly by the BLS
and the USCB to assess workforce characteristics, the likelihood of a person in the United States to be employed as a PT
was averaged during the past 10 years in the following age
groups: 16 to 19, 20 to 24, 25 to 34, 35 to 44, 45 to 54, 55 to
64, and 65 years and older. These 10-year proportions were
multiplied by USCB-projected population increases in each
age category, providing an estimate of the number of persons
in each age group that will assume a PT job each year until
2030. This estimate was summed across age groups for each
year through 2030 and added to the 2008 BLS-reported PT
supply, generating estimates of PT supply through the year
2030. The formula below contains details of the supply
model,
SR, 2009 ⫽ BLS2008 ⫹
SR, N ⫽ BLS2008 ⫹
⫹
兺
R
兺
R
兺
R
(LA ⫻ [⌬POPA, 2008, 2009])
(LA ⫻ [⌬POPA, 2008, 2009])
(LA ⫻ [⌬POPA, 2009, 2010]) ⫹ . . .
⫹
兺
R
(LA ⫻ [⌬POPA, N-1, N])
where SR,N S ⫽ supply, R ⫽ region or State, and N ⫽ year; LA:
L ⫽ likelihood averaged over 10 years, A ⫽ age group;
⌬POPA, N-1,N ⫽ age-category–specific populationN – populationN-1; and BLS2008 ⫽ the number of PT jobs reported by
the BLS in 2008.
Report Card
The metric used for grading in this work was the PT shortage
ratio, which is defined as the difference between PT demand
and PT supply per 10,000 persons and demonstrated by the
following equation:
[State] PT Demand ⫺ [State] PT Supply
⫻ 104
[State] Total Population
⫽ [State] PT Shortage Ratio
Population statistics were obtained from the USCB [22].
The national and state number of PT jobs in 2008 were
obtained from the BLS [24]. The national PT supply ratio, or
national mean, 5.50 PT jobs per 10,000, served as the target
or “demand” value as well as the “C” grade in the rubric. The
standard deviation of the PT supply ratios across the 50 states
formed the framework of the grading rubric. A and F grades
were ⫾2 standard deviations, B and D grades were ⫾1
standard deviation, and C⫹ and C⫺ grades were ⫾0.5
standard deviation (Table 3).
As discussed previously, shortages were determined as the
difference between projected PT demand and PT supply.
Letter Grade
Shortage Grading Rubric
A
B
C⫹
C
C⫺
D
F
ⱕ ⫺2.99
⫺2.98 to ⫺1.49
⫺1.48 to ⫺0.75
⫺0.74 to 0.74
0.75 to 1.50
1.49 to 2.98
ⱖ 2.99
The ranges are PT shortage ratios (developed from 2008 data of Bureau of
Labor Statistics). A negative ratio implies a scenario in which supply exceeds
demand.
This figure was divided by the projected population for each
region and each year, yielding a PT shortage ratio through the
year 2030. With the use of the grading rubric described
previously, grades were assigned as a measure of each state’s
PT shortage ratio for the years 2008, 2020, and 2030. These
dates were chosen in absence of other models for comparison
to provide current, interval, and the furthest predicted shortage ratio projection possible.
The upper and lower estimates of PT shortages (the number of PT jobs) were generated for each state. The standard
error of the demand slope model was used to generate 95%
confidence intervals for each state. Similarly, the standard
error of the 10-year mean likelihood values calculated for
each age category in the supply model were used to generate
95% confidence intervals for each age category, each year,
and each state. The upper and lower limits of each confidence
interval were summed to yield high and low estimates, respectively, of PT supply. The lower limit of the shortage range
was generated by calculating the difference between lower
demand estimates and upper supply estimates. Similarly, the
shortage range upper limit was produced by calculating the
difference between upper demand estimates and lower supply estimates.
RESULTS
Our results demonstrate that there are PT shortages in many
states, and these shortages are forecasted to increase through
the year 2030. Figure 1 displays the assigned shortage ratio
grades and PT shortages (the number of PT jobs) for each
state [17]. Our results demonstrate that in 2008, 12 states
exhibited PT shortage ratio grades with C⫺ or lower. By the
year 2020, 34 states will exhibit PT shortage ratio grades with
C⫺ or lower; and by the year 2030, 48 states will exhibit PT
shortage ratio grades with C⫺ or lower. In the year 2008, the
states with the highest shortage ratios were in the southern
and western regions and included the following states: Alabama (grade D), Georgia (grade D), Nevada (grade D), and
California (grade D). Our model forecasts that in 2030, the
states with the highest shortage ratios will be South Carolina
(5.03, grade F), Georgia (5.16, grade F), California (5.34,
PM&R
Vol. 2, Iss. 11, 2010
1025
Our model forecasts of PT supply ratios and PT demand
ratios are graphically displayed in Figure 2 for each of the
BLS-defined regions. These data demonstrate that the PT
supply ratios will remain relatively constant, whereas the PT
demand ratios will increase, outpacing supply and creating
shortages. This finding is especially true for the southern and
western regions that currently exhibit shortages.
The PT shortage (the number of PT jobs) from 2008
through 2030 for each BLS-defined region is displayed in
Figure 3. It can be seen from Figure 3 that our model
forecasts an increase in PT shortages in all regions and states,
with the southern and western regions exhibiting the largest
increases in PT shortages.
DISCUSSION
Figure 1. Assigned shortage ratio letter grades for the nation
(50 states) and each state for the years 2008, 2020, and 2030
and PT shortages (the number of PT jobs) calculated for each
state for 2008 and 2030. Negative numbers imply a scenario in
which supply exceeds demand on the basis of our models.
States are grouped by the BLS US regional groupings [17].
grade F), New Mexico (6.39, grade F), Florida (6.81, grade
F), Nevada (6.95, grade F), and Arizona (7.53, grade F). By
2030, the states with the lowest shortage ratios, and the greatest
supply of PTs relative to demand, will be in the northeastern
region and include Massachusetts (⫺0.65, grade C) and Connecticut (⫺0.60, grade C). In 2008, there were 3 states with
letter grade A, 7 Bs, 10 C⫹s, 18 Cs, 8 C⫺s, 4 Ds, and 0 Fs. By the
year 2030, our model forecasts 0 As, 0 Bs, 0 C⫹s, 2 Cs, 3 C⫺s,
18 Ds and 27 Fs. By using a numeric grading scale in which A ⫽
4, B ⫽ 3, C⫹ ⫽ 2.33, C ⫽ 2, C⫺ ⫽ 1.67, D ⫽ 1, and F ⫽ 0, we
found that in 2008 the average national grade was 2.19. By
the year 2030, we forecast the average national grade to be
0.54. Similarly, by 2030, the following states will demonstrate the largest PT shortage (the number of PT jobs): California (24,782 jobs), Florida (19,541 jobs), and Texas
(16,346 jobs).
Our model forecasts that, on the basis of current trends,
demand for PT services will far outpace the supply of PTs
within the U.S. According to our model, by the year 2030,
most states will have severe PT shortages. These shortages
will be most severe in the western and southern regions. This
is a very sobering forecast; the cost and expenses associated
with training therapists to meet the demand are in the range
of billions of dollars. Thus, it requires concerted and persistent efforts in all segments of society, government at all levels,
the health care industry, professional societies, academia,
and the public at large to put forth sustainable growth plans
for meeting the challenges of tomorrow.
Examination of the health care workforce is difficult because of the complexity of factors that affect supply, demand,
and balance between supply and demand. Health-care policy
is one of these factors. The supply-and-demand models used
in this study are determined by health care practice regulated
by current policies. Using an approach similar to ours, the
APTA Vector Report predicted a PT surplus by the year 2005
[9] that did not appear to occur. The failure of this prediction
is attributable in part to changes in health care policy (eg,
institution of and then moratoriums delaying implementation of the Medicare cap imposed by the BBA). If health care
policy had remained the same from 1997 to 2005, the Vector
Report surplus prediction may have been realized. One key
provision of the health care reform bill recently passed by
Congress is to insure the uninsured, which presumably will
increase access to health care services, increasing demand
beyond current predictions.
To evaluate the effects of health care reform and other
health care policies, strategies need to be developed to effectively monitor supply and demand of the health care workforce. These strategies should be at the state and national
levels, be ongoing, and be collaborative among key stakeholders. For example, Texas nursing leadership worked with
key stakeholders to pass legislation addressing the nursing
shortage in Texas [25]. This collaborative effort led to development of a statewide health council focused on health
1026
Zimbelman et al
PHYSICAL THERAPY WORKFORCE IN THE U.S.
Figure 2. Projected PT supply ratio and PT demand ratio for each aggregate BLS-defined region (see Table 2). Shaded regions
indicate years in which PT supply ratio meets or exceeds PT demand ratio according to our model forecasts.
professional workforce planning. This council serves as a
resource to develop legislation and facilitate projects to ensure that health care services and resources are available
throughout the state of Texas [25]. To determine the supply
and distribution of health care workers, the state of North
Carolina established a comprehensive and continuously
maintained database (North Carolina Health Professions
Data System) that tracks supply and demand of licensed
health professionals within the state [26]. Reports released
from these data include historical examination of supply
ratios throughout the state reporting distribution disparities
in counties, urban and rural areas, vacancy rates, and recommendations for policy makers and educators [26-28]. The
work in Texas and North Carolina are examples of health care
workforce planning that could be adopted by states throughout the United States to establish state and federal policies to
ensure adequate supply of health professionals.
State and federal policies may also be used to mitigate
regional disparities in the distribution of PTs. In the current
study, all states that received a PT shortage ratio grade of C or
lower had population densities of less than 448 persons per
square mile, and 27 of these 30 states had population densities of less than 147 persons per square mile [29]. For
example, Massachusetts and Connecticut had low shortage
ratios with high population densities (833 and 723 persons
per square mile, respectively), whereas Nevada and Alabama
had high shortage ratios with low population densities (24
and 92, respectively). Although this trend was not true for all
states (eg, Vermont had a low shortage ratio and low population density, 67 persons per square mile), in general, the
states that exhibited current PT shortages represented more
rural areas.
The authors of previous studies have demonstrated similar distribution disparities across the nation [10,11] and
PM&R
Figure 3. PT shortages (the number of PT jobs) projected for
each BLS-defined region (see Table 2). Negative numbers
imply a surplus of PT jobs according to our model.
across counties [26,27,30]. PTs tend to practice in more
populated areas [30], in/near cities and larger rural towns
[26]. This distribution disparity is related more to population
size and the availability of health facilities in which to practice
and less to economic factors [30]. The geographic and rural/
urban distribution disparities are not unique to PT but are
reported in other health professions [5,26,31] and in other
countries [5,32,33], suggesting that people living in rural
areas are being underserved by the health care workforce.
Programs exist to address health care workforce distribution disparities for health professions other than PT. For
example, the National Health Service Corps is a federal
program that provides loan repayment and scholarships for
primary care providers (physicians, nurse practitioners, certified midwives, physician assistants, dentists, dental hygienists, and mental health workers) who commit to work in
Health Professional Shortage Areas [34]. Students at the
University of Minnesota are immersed in rural medical practice for 36 weeks; this program has resulted in successfully
graduating rural primary care physicians [35]. The state of
Wisconsin established a consortium to study the misdistribution of primary care providers within the state [36]. The
Wisconsin consortium developed recommendations resulting in health care education program changes, development
of pilot programs designed to meet the health care needs of
underserved populations, and the restructuring of loan repayment programs designed to attract primary health care
providers to underserved regions. Currently, none of these
programs include PT. Implementation and expansion of
these types of programs to include PT would serve to reduce
and eventually eliminate the misdistribution of PTs, ensuring
that the health care needs of the underserved are met.
The problem of health care workforce shortages is not
unique to PT but has been observed in other health professions (eg, nursing [15,16,37-39] and medicine [40-42]).
Vol. 2, Iss. 11, 2010
1027
Solutions to mitigate PT shortages may be found by examining what these other professions have done. For example,
evidence linking hospital nurse staffing to patient outcomes
and improving quality and safety of patient care has demonstrated to stakeholders that adequate numbers of well-trained
nurses will aid in achieving higher quality health care delivery systems [39]. Developing standard outcome measures of
PT, providing evidence that PT is critical to high-quality
health care delivery, and educating stakeholders will
strengthen the profession. Improving the workplace to aid
retention of PTs is also critical to mitigating shortages. For
nursing, this has been accomplished in part by instituting the
Magnet Hospital Recognition Program (American Nurses
Credentialing Center) in which organizations are recognized
for excellence in nursing practice and patient care [39]. This
has improved job and career satisfaction among nurses, aiding retention efforts [39]. The APTA has examined PT turnover rates in 3 practice settings [12-14] and reported greater
turnover rates in acute care hospitals and skilled nursing
facilities than in outpatient private practices. Identifying the
reasons for high turnover rates would be prudent to develop
initiatives aimed at retention of skilled practitioners in these
practice settings.
Another mechanism for mitigating shortages may be
through foundations and private sector initiatives. The private sector could provide resources to test initiatives aimed at
improving the workplace, conduct studies on quality of PT
services, and support PT education. Private initiatives could
aid in increasing public awareness of the PT shortages, convey positive images of the PT profession to the public, and
raise funds for PT education. This was accomplished for
nursing and resulted in strengthening the nursing profession
and reducing shortages [39]. Increasing access to health care
occupation education through e-learning technologies, better collaborations between clinical sites and educational programs, and creative public/private partnerships as demonstrated in Iowa [43] and Oregon [44] will also help mitigate
health care workforce shortages.
Limitations/Assumptions
There are several limitations and assumptions inherent in our
models. First, the demand model is determined only by age
and population growth, assumes that growth is linear, and
assumes that it is the same across all states. Although this
assumption is simplistic, it is a first attempt to use available,
creditable data to examine both supply and demand for the
entire country. Although population size and age are fundamental components of health care utilization in a number of
national workforce projections [45], health insurance status,
regional variations in service costs, and even regional industry composition all play roles in determining PT demand. We
did not account for these other factors in our demand model.
Another limitation of the demand model is that it assumes a
1028
Zimbelman et al
static health care economic and policy structure. Because
health care coverage is linked to employment, fluctuations in
the economy that affect employment will affect PT demand.
Medicare and Medicaid policies, established by federal and
state policy makers, effects access to PT and reimbursement
for PT services. As demonstrated with the BBA, the number of
PT jobs is greatly influenced by changes in health care policy.
Health care reform policies recently passed by Congress and
the current economic recession have the potential to significantly affect our model predictions. With the increasing
focus on health and wellness and prevention, the scope of PT
practice has changed, creating additional demand for PT
services in other settings such as health clubs, professional
athletic organizations, and senior centers that did not exist
previously [46,47]. Direct access to PT (instead of via referral) has the potential to increase demand because access to PT
services by the general public would be easier. Another factor
affecting our demand model is the use of 1999 PHE from the
Centers for Medicare & Medicaid Services. Older age groups
tend to have greater PHEs than younger age groups because
of the development of chronic conditions associated with
aging [21]. It has been reported that chronic disability is
decreasing [48]. Reducing the prevalence of chronic conditions may impact PHE and may affect our demand projections.
Our supply model also incorporates a number of key
assumptions that may limit its conclusions. The primary
assumption is that PT job supply is driven by age, population
growth, and historic likelihood of specific age groups to be
employed as a PT. It does not incorporate workplace settings,
part-time or full-time employment status, the impact of gender or ethnic factors on PT jobs, interstate migration of PTs,
international recruitment of PTs, or state and national economic and political factors that impact desirability of PT
employment. For example, the number of applicants to PT
programs declined from 1995 to 2002 [49], possibly because
of the reduction of PT jobs after the BBA was passed. Since
2002, the number of applicants to PT programs has been
increasing [49], potentially because of continued reports that
PT is a desirable profession.
Finally, our grading system is determined via the current
national PT supply ratio of 5.50. We assigned a letter grade of
C to this ratio because this was the national ratio in 2008.
Anything above a letter C was considered to be a surplus (or
at least less of a shortage); a letter grade of C would suggest a
balance between PT supply and PT demand. This is most
likely not the case, given the many job postings across the
nation, sign-on bonus incentives, and the 2008 APTA workforce study reporting significant proportions of job vacancies
[12-14].
CONCLUSION
The forecasts in this study suggest that shortages of PTs
currently exist in many states across the nation and that these
PHYSICAL THERAPY WORKFORCE IN THE U.S.
shortages will increase over the next 2 decades. These results
should be used as points of discussion for policy makers and
stakeholders to examine the issue of PT supply and PT
demand. Follow-up studies need to be performed to validate
our model, confirm our current results, and develop more
complex models that incorporate other factors affecting supply and demand not accounted for in our model. It is important that adequate models of supply and demand for the PT
profession be developed. These models should be used in
ongoing assessment of PT distribution to identify underserved regions, practice settings, and populations. In addition, ongoing determination of shortages/surpluses needs to
occur in order to adequately forecast future trends. Monitoring these trends over time is critical for planning and policy
making to ameliorate severe shortages or surpluses and maintain an appropriate balance between PT supply and PT demand that meets population needs.
ACKNOWLEDGMENTS
The authors thank Amanda Chiu for assistance with data
processing, Walter Liau and Daniel Chen for coordinating
the research program, and Vinoth Ranganathan for editorial
comments and assistance.
REFERENCES
1. U.S. News and World Report, Best Careers 2009. 2008. Available at:
http://www.usnews.com/articles/business/best-careers/2008/12/11/
best-careers-2009-physical-therapist.html?PageNr⫽1. Accessed January 16, 2009.
2. Bureau of Labor Statistics, Occupational Outlook Handbook, 2008-09
Edition. 2008. Available at: http://www.bls.gov/oco/ocos080.htm. Accessed January 16, 2009.
3. Hack LM, Konrad TR. Determination of supply and requirements in
physical therapy: Some considerations and examples. Phys Ther 1995;
75:47-53.
4. Dreesch N, Dolea C, Dal Poz MR, et al. An approach to estimating
human resource requirements to achieve the Millennium Development
Goals. Health Policy Plan 2005;20:267-276.
5. Zurn P, Dal Poz MR, Stilwell B, Adams O. Imbalance in the health
workforce. Hum Resour Health 2004;2:13.
6. Selker LG. Human resources in physical therapy: Opportunities for
service in a rapidly changing health system. Phys Ther 1995;75:31-37.
7. Jacoby I. Forecasting requirements for physical therapists. Phys Ther
1995;75:38-44.
8. APTA Employment Survey. Alexandria, VA; American Physical Therapy
Association. 2005. Available at: http://www.apta.org/AM/
Template.cfm?Section⫽Surveys_and_Stats1&Template⫽/TaggedPage/
TaggedPageDisplay.cfm&TPLID⫽149&ContentID⫽18996. Accessed
September 4, 2009.
9. Vector Research Inc. American Physical Therapy Association Supply
and Requirements Model Findings.University Place, WA: Vector Reserach Incorporated; 1997.
10. Chevan J, Chevan A. A statistical profile of physical therapists, 1980
and 1990. Phys Ther 1998;78:301-312.
11. Landry MD, Ricketts TC, Fraher E, Verrier MC. Physical therapy health
human resource ratios: A comparative analysis of the United States and
Canada. Phys Ther 2009;89:149-161.
PM&R
12. Physical Therapy Workforce Project. Physical Therapy Vacancy and
Turnover Rates in Outpatient Private Practices. Alexandria, VA: American Physical Therapy Association; 2008.
13. Physical Therapy Workforce Project. Physical Therapy Vacancy and
Turnover Rates in Acute Care Hospitals. Alexandria, VA: American
Physical Therapy Association; 2008.
14. Physical Therapy Workforce Project, Physical Therapy Vacancy and
Turnover Rates in Skilled Nursing Facilities. Alexandria, VA: American
Physical Therapy Association; 2008.
15. Lin VW, Lee A, Juraschek S, Jones D. California regional registered
nurse workforce report card. Nurs Econ 2006;24:290-297, 279.
16. Lin VW, Juraschek SP, Xu L, Jones D, Turek J. California regional
registered nurse workforce forecast. Nurs Econ 2008;26:85-105, 121.
17. Bureau of Labor Statistics. How are “employees” defined by the OES
survey. 2010. Available at: http://www.bls.gov/oes/oes_ques.htm#
Ques16. Accessed May 2, 2010.
18. Bureau of Labor Statistics. Standard Occupational Classification. 2010.
Available at: http://www.bls.gov/soc/. Accessed July 13, 2010.
19. U.S. Census Bureau. Population Estimates: Methodology. 2010. Available at: http://www.census.gov/popest/topics/methodology/. Accessed
May 2, 2010.
20. Bureau of Labor Statistics, Region Definitions. 2008; http://www.bls.
gov/eag/home.htm. Accessed August 1, 2008.
21. Keehan S, Lanzenby H, Zezza M, Catlin A. Age Estimates in the
National Health Accounts. Baltimore, MD: Centers for Medicaid and
Medicare Services; 2004.
22. United States Census Bureau (USCB). Population Division, Interim
State Population Projections by Single Year of Age: July 1, 2004 to
2030. 2008. Available at: http://www.census.gov/population/www/
projections/projectionsagesex.html. Accessed August 1, 2008.
23. U.S. Census Bureau Current Population Survey (CPS): A Joint Effort
Between the Bureau of Labor Statistics and the Census Bureau 20032007. 2008. Available at: http://www.census.gov/cps/. Accessed August 1, 2008.
24. Bureau of Labor Statistics (BLS), Healthcare Practitioner and Technical
Occupations Profiles, 2006. 2008. Available at: http://www.bls.gov/
oes/current/oes_stru.htm#29-0000. Accessed August 1, 2008.
25. Kishi A, Green A. A statewide strategy for nursing workforce development through partnerships in Texas. Policy Polit Nurs Pract 2008;9:
210-214.
26. Gaul KFE, Guild P, Ricketts T, Hadley H, Darter J. Trends in Licensed
Health Professionals in North Carolina: 1979-2005. Chapel Hill, NC:
Cecil G. Sheps Center for Health Services Research, University of North
Carolina-Chapel Hill; 2007.
27. Panel NCPTWAT.Maintaining Balance: The Physical Therapy Workforce in North Carolina in the Year 2000. Chapel Hill, NC: Cecil G.
Sheps Center for Health Services Research, University of North Carolina-Chapel Hill; 2000.
28. Thaker SFE, King J. Allied Health Job Vacancy Tracking Report. Chapel
Hill, NC: Cecil G. Sheps Center for Health Services Research, University of North Carolina-Chapel Hill; 2006. Available at: http://www.
shepscenter.unc.edu/hp/publications/AHvacancy_spring06.pdf. Accessed April 23, 2009.
29. U.S. Census Bureau Population Estimates: Persons per Square Mile.
2008. Available at: http://factfinder.census.gov/jsp/saff/SAFFInfo.jsp?_
pageId⫽thematicmaps&_submenuId⫽maps_1. Accessed July 23, 2009.
Vol. 2, Iss. 11, 2010
1029
30. Wilson SD, Langwell KM, Deane RT, Chui KF, Black RA. Identification
of physical therapist shortage areas: A study of thirteen states. Phys
Ther 1982;62:315-323.
31. Wilson RD, Lewis SA, Murray PK. Trends in the rehabilitation therapist
workforce in underserved areas: 1980-2000. J Rural Health 2009;25:
26-32.
32. Baumann A, Hunsberger M, Blythe J, Crea M. Sustainability of the
workforce: Government policies and the rural fit. Health Policy 2008;
85:372-379.
33. Scheffler RM, Liu JX, Kinfu Y, Dal Poz MR. Forecasting the global
shortage of physicians: An economic- and needs-based approach. Bull
World Health Organ 2008;86:516-523B.
34. Health Resources and Services Administration. 2009. Available at:
http://bhpr.hrsa.gov/shortage/. Accessed August 3, 2009.
35. Zink T, Halaas GW, Finstad D, Brooks KD. The rural physician associate program: The value of immersion learning for third-year medical
students. J Rural Health 2008;24:353-359.
36. Riportella-Muller R, Littlefield V, Libby DL. Nursing drives interdisciplinary health care workforce planning in Wisconsin: One state’s
experience. J Prof Nurs 2000;16:169-176.
37. Kovner CT, Mezey M, Harrington C. Who cares for older adults?
Workforce implications of an aging society. Health Aff (Millwood)
2002;21:78-89.
38. Unruh LY, Fottler MD. Projections and trends in RN supply: What do
they tell us about the nursing shortage? Policy Polit Nurs Pract 2005;
6:171-182.
39. Buerhaus PI. Current and future state of the US nursing workforce.
JAMA 2008;300:2422-2424.
40. Deal CL, Hooker R, Harrington T et al. The United States rheumatology
workforce: Supply and demand, 2005-2025. Arthritis Rheum 2007;56:
722-729.
41. Williams TE, Jr., Ellison EC. Population analysis predicts a future
critical shortage of general surgeons. Surgery 2008;144:548-554.
42. Kimball AB, Resneck JS Jr. The US dermatology workforce: A specialty
remains in shortage. J Am Acad Dermatol 2008;59:741-745.
43. Bristol TJ, Gloor EM. Development of a strategic plan to increase access
to health occupations education in Iowa. Nurs Educ Perspect 2007;28:
337-341.
44. Oregon Simulation Alliance. 2008. Available at: http://www.
oregonsimulation.com/content/about-osa. Accessed September 14,
2009.
45. Health Resources and Services Administration, Projected Supply, Demand, and Shortages of Registered Nurses: 2000-2020. 2010. Available
at: http://bhpr.hrsa.gov/healthworkforce/reports/behindrnprojections/
3.htm. Accessed May 2, 2010.
46. Loria K. Rink ready. PT Magazine Phys Ther 2009;17:24-28.
47. Wojciechowski M. Expanding into new practice areas. PT Magazine
Phys Ther 2009;17:18-22.
48. Manton KG. Recent declines in chronic disability in the elderly U.S.
population: Risk factors and future dynamics. Annu Rev Public Health
2008;29:91-113.
49. Physical Therapist Education Programs 2007-2008 Fact Sheet. May
2008. Alexandria, VA: American Physical Therapy Association;
2008.