Keywords
resident education - cost of training - housestaff - cost of education
Medical training has historically been an apprenticeship, variably balancing service
and education. Despite increasing regulation and oversight, the argument of whether
a resident physician is a student or employee remains unresolved. On one hand, the
National Labor Relations Board in 1999 ruled residents are employees under the National
Labor Relations Act.[1] On the other hand, Graduate Medical Education (GME) has been subsidized since the
signing of Titles XVII and XIX of the Social Security Act in 1965 and the establishment
of Medicare and Medicaid.[2] In fiscal year 2012, Medicare provided approximately $9.7 billion in funding to
hospitals for GME, with more than two-thirds ($6.8 billion) designated as indirect
payments. Intended to help defray the assumed and uncalculated increased costs of
providing patient care in a teaching hospital, these indirect payments have repeatedly
been the target of budget cuts by regulatory bodies.[3]
[4] In fact, a 2014 report from the Institute of Medicine recommended replacing the
current payment model for GME, noting “remarkably little is known about the individual,
institutional, and societal costs of residency training … Federal GME regulations
are nearly silent regarding transparency and accountability for use of Medicare GME
funds.”[5]
Federal funding cuts for teaching hospitals have understandably been met with concern;
a national survey of GME directors suggested that even a modest decrease in indirect
Medicare reimbursement would likely trigger downsizing of impacted programs.[6] Since the training of residents presumably leads to increased hospital costs, replacing
or augmenting housestaff with advanced practice providers (APPs) is one potential
response to projected budgetary constraints.[7]
[8] We performed a direct cost–benefit analysis of this theoretical approach in the
Department of Ophthalmology at the University of Kentucky, evaluating a small component
of the cost structure for the particular types of personnel involved in delivering
care.
Materials and Methods
To determine cost, the total mean salary and benefits of all APPs (nurse practitioners,
clinical nurse specialists, nurse anesthetists, and physician assistants), optometrists,
faculty ophthalmologists, and ophthalmology residents at the University of Kentucky
were determined utilizing available financial data for fiscal year 2017. The resident
salary and benefits represented the mean for the complement of ophthalmology residents.
APPs, optometrists, and faculty are salaried and the salary and benefits represent
the mean for all APPs in the College of Medicine and all optometrists and faculty
ophthalmologists in the Department of Ophthalmology. The cost per week based on mean
number of hours worked was calculated for each group, as determined by the mean number
of hours worked. The number of hours worked/week for residents was the mean of 2016
to 2017 duty hour logs per Accreditation Council for Graduate Medical Education (ACGME)
requirements; for APPs, it was based upon institutional fiscal calculations.
To determine benefit, the mean number of work relative value units (wRVUs), technical
collections, and expenses for the three groups during fiscal year 2017 were tabulated.
For optometrists and faculty ophthalmologists, these figures were adjusted for clinical
full-time equivalents. Net revenue was technical collections subtracted from total
expenses.
Results
Total annual mean salary and benefits for 242 APPs, 4 optometrists, 20 faculty ophthalmologists,
and 9 ophthalmology residents were $126,797, $117,021, $338,233, and $71,210, respectively
([Table 1]). APPs worked a maximum of 50 hours/week and ophthalmology residents worked an average
of 60 hours/week. Given the variability in work hours across the four groups, an average
of 50 hours/week was selected to provide comparable analysis. On that basis, the calculated
hourly cost was $48.77, $45.01, $130.09, and $27.39, respectively ([Table 1]).
Table 1
Mean total salary, benefits and cost/hour of advanced practice providers, optometrists,
faculty ophthalmologists, and ophthalmology residents
|
Salary and benefits
|
Cost/hour
|
|
Position
|
Salary
|
Total benefits
|
Total (salary + benefits)
|
40 h/wk
|
50 h/wk
|
60 h/wk
|
|
APP
|
$98,456
|
$28,341.16
|
$126,797
|
$60.96
|
$48.77
|
$40.64
|
|
Optometrist
|
$89,500
|
$27,520.77
|
$117,021
|
$56.26
|
$45.01
|
$37.51
|
|
Ophthalmology faculty
|
$268,440
|
$69,794.15
|
$338,233
|
$162.61
|
$130.09
|
$108.41
|
|
Ophthalmology resident
|
$54,570
|
$16,639.97
|
$71,210
|
$34.24
|
$27.39
|
$22.82
|
Abbreviation: APP, advanced practice provider.
Mean number of hours worked for each position.
Total annual mean wRVUs and technical collections are presented in [Table 2]. Ophthalmology residents do not directly generate wRVUs or collections. Accordingly,
the net annual revenue was−$62,729, $122,757, $566,119, and −$71,210, respectively.
Table 2
Mean relative collections, expenses, and net revenue for advanced practice providers,
optometrists, faculty ophthalmologists, and ophthalmology residents
|
APP
|
Optometrist[a]
|
Ophthalmology faculty[a]
|
Ophthalmology resident
|
|
wRVU
|
1,477
|
5,234
|
10,991
|
0
|
|
Charges
|
$222,622
|
$701,467
|
$2,667,194
|
$0
|
|
Payments
|
$64,048
|
$239,778
|
$904,352
|
$0
|
|
Technical collections
|
$64,048
|
$239,778
|
$904,352
|
$0
|
|
Total expenses
|
$126,777.00
|
$117,021
|
$338,233
|
$71,210
|
|
Net income
|
($62,729)
|
$122,757
|
$566,119
|
($71,210)
|
Abbreviations: APP, advanced practice provider; wRVU, work relative value unit.
a Adjusted for clinical full-time equivalents.
Discussion
A report by the RAND corporation in 2013 evaluating the financial impact of residency
training programs concluded, “[i]f the hospital has service needs that would otherwise
need to be met by hiring alternative providers, there is a marginal benefit to adding
a resident, particularly in a subspecialty program, before considering the additional
benefits of any GME-related revenues.”[7] A more recent analysis at the University of Massachusetts Medical School examining
the opposite—the financial implications of reducing residency programs—found that
decreasing the size of large programs results in higher replacement costs than the
direct costs of residents, and the marginal gains from eliminating small programs
are far smaller than the estimated shortfall if Medicare GME funding is reduced.[8] Our study augments these results by demonstrating the significantly lower cost of
a resident compared with a replacement APP, optometrist, or faculty ophthalmologist.
At our institution, the hourly cost of an ophthalmology resident is 40, 44, and 23%
of that of an APP, optometrist, or faculty ophthalmologist. However, because resident
physicians cannot bill for services, the direct revenue was notably disparate, with
residents and APPs both generating net negative income compared with optometrists
and faculty ophthalmologists. For the APPs, the negative income is mitigated by the
increased clinical, hospital, and surgical productivity of physicians, and these projections
are incorporated into overall cost analyses of each of these positions. Similar data
are not currently available for resident physicians.
The number of APPs is growing at a rate greater than physicians, largely in response
to the increasing divide between supply and demand in health care. In the field of
ophthalmology, however, this has yet to occur; in 2015, there were reportedly only
70 physician assistants (PA) employed by ophthalmologists in the United States.[9] Potential solutions proposed to address the demand include enhancing the scope of
optometry, expanding the number of residency positions, increasing the workload of
ophthalmologists, and/or utilizing APPs.[10] A direct comparison of these employee groups is difficult, given an inclusive substitute
for the clinical and surgical responsibilities of our trainees is not currently available.
One very recent report attempting to bridge the gap from the Wilmer Eye Institute
discussed outcomes after incorporating a PA into their consult service. The PA was
a former comprehensive ophthalmologist in another country and works for 3.5 days a
week, served as the primary responder for all consults, calling the second-year resident
on rotation only for those patients requiring surgical intervention, were of educational
interest, had abnormal findings or required subspecialty consultation. The authors
found most residents strongly agreed that having a PA both improved education and
the balance between service and education. The PA has been cost neutral to the department,
and the authors estimate the break-even point in their department is roughly 8 to
12 patient visits per day.[11]
Since resident physicians cannot independently bill for clinic visits or operative
cases, they do not directly generate revenue, as demonstrated by the wide gap between
costs and benefits in this analysis. Limited data and transparency surrounding trainee
impact on hospital productivity are part of the reason GME funding is continually
subject to scrutinization.[12] As this analysis demonstrates from a direct cost comparison perspective, resident
physicians are significantly cheaper sources of labor compared with any potential
replacements. While it is not possible to calculate the indirect revenue generated
by resident physicians, two studies of general surgery residents at single institutions
estimated the unbilled revenue of a resident was $95,000[13] or $233,000[14] per year. A more recent multicenter prospective analysis determined unbilled consult
services provided by orthopedic residents during on-call hours would fund 73% of the
resident's stipends.[15] Conversely, and as stated above, a calculation of indirect costs of resident physicians
is currently problematic. While it is assumed that training a physician increases
the time and costs of health care, there are limited data to support this.[16] As institutions grapple with the overall financial impact of current and additional
APPs, similar data should be collected and analyzed for our trainees.
Overlying any discussion on the role of resident physicians in the health care labor
force is the perceived dichotomy between education and service, a fundamental oversight
function of the ACGME. While the definition of service is necessarily imprecise, and
service obligations can serve an educational role,[17] any sincere effort to evaluate the utility of resident physicians in this context
will need to consider educational objectives and well-being.
There are several important limitations to this analysis. It represents a single department
at one academic medical center and it assumes that APP, optometry and ophthalmology
faculty wRVU, and technical collections would remain unchanged if their job responsibilities
mirrored the work of a resident physician. Further, residency programs and training
environments are diverse, with university, veteran affairs, and community-based clinics
as well as county or indigent locations. Programs that offer charity services in resident-run
clinics are an important example of potential costs and benefits outside the scope
of this analysis. Like most ophthalmology practices,[10] we currently have no APPs employed in our department, so the comparison does not
directly relate to ophthalmology-extenders. As stated above, we cannot accurately
calculate any indirect costs or benefits. Lastly, this analysis focuses solely on
the comparative financial cost and benefits of these providers. It does not address
access to or quality of care or any societal need to train future care providers.
As Albert Einstein is attributed with saying, “Not everything that can be counted
counts, and not everything that counts can be counted.”
In summary, we found resident physicians to be relatively inexpensive providers compared
with APPs, ophthalmology faculty, or optometrists at our medical center. Because trainees
are unable to bill for services, the net revenue was lower than the other groups.
A comprehensive response to both predicted and unforeseen changes in health care will
likely utilize a combination of options, but our data suggest that in academic centers,
resident physicians are a relatively inexpensive source of labor compared with optometrists,
faculty ophthalmologists, and APPs. As stakeholders continue to evaluate the landscape
of health care and the role of different provider models in addressing evolving demands,
we need to incorporate a frank and transparent discussion surrounding the role and
financial implications of trainee involvement in care.
