Academic Psychiatry 24:195-201, September 2000
© 2000 Academic Psychiatry
The Cost of Residency Training in Psychiatry
Jed Magen, D.O. and
Deborah Banazak, D.O.
Dr. Magen is Director of Residency Education and Assistant Professor, Department of Psychiatry, Michigan State University. Dr. Banazak is Associate Director of Residency Education and Assistant Professor, Department of Psychiatry, Michigan State University. Address reprint requests to Dr. Magen, Michigan State University, Department of Psychiatry, A-229 East Fee Hall, East Lansing, MI 48824-1316. e-mail: magenj{at}msu.edu
|
ABSTRACT
|
Graduate medical education (GME) in the United States is financed largely by the federal government through the Medicare program. Medicare regulations developed in 1997 resulted in cuts in the direct, indirect, and disproportionate-share components of GME funding. Given this situation, it is increasingly important for programs to understand what the true costs of residency training are and how they are allocated. Using a faculty time survey and the yearly residency budget, the authors report on such an analysis for their own training program (at Michigan State University) for the 1996–97 academic year. Overall, the response rate for the faculty time survey was 77% (N=28). The identified program costs of supporting 20 psychiatry residents, excluding donated time, were calculated at $1,563,193, or $78,159 per resident. Eighty-four percent of costs were for personnel. These costs are consistent with those derived from studies of family practice and internal medicine training programs. Awareness of the costs of training will assist training directors in budget negotiations with their institutions.
Key Words: Residency Training • Training Costs
|
INTRODUCTION
|
Graduate medical education (GME) in the United States is financed largely by the federal government through the Medicare program. Some hospitals (primarily academic medical centers) derive large amounts of per-resident funding from the direct and indirect components of GME funding (1,2). However, controversy continues around the true cost of GME to hospitals. The Balanced Budget Act of 1997 resulted in cuts in the direct, indirect, and disproportionate-share components of GME funding (3,4). Such cuts are adversely affecting academic medical centers, which already have difficulty supporting investment in new technologies and programs (5). Other revenues, such as patient revenues, that have traditionally been cost-shifted to support training, are also falling (6). Academic practice plans, which, in 1997–98, comprised 33.9% of medical school revenues, may be stretched even further (7,8). Most managed-care organizations refuse to pay for resident training and will not reimburse for trainee time. Caps on residency numbers tend to turn training slots into a commodity, with perhaps greater value placed on training programs that are more remunerative to hospitals. One can guess that less remunerative residencies, such as psychiatry, may be affected to a larger extent than programs in disciplines that generate sizeable revenues for hospitals and therefore may be more highly valued.
Most psychiatry residencies are under increasing pressure; others have combined; and several have closed. Hospitals have a finite GME revenue stream that must be allocated to various programs. In this environment, knowledge is power. Chairs and program directors who understand their program costs will better understand where to make cuts, if necessary; they can advocate for program funding by using this information, and they may be able to detect synergies and areas where cooperation with other programs may be cost-effective.
A computerized MEDLINE search (1960–present) failed to uncover relevant literature that directly measures costs associated with psychiatry training. Only a small literature, all of which details the costs associated with family practice and internal medicine training programs, exists. As is often the case, many studies are not easily comparable because different methodologies are used to allocate and calculate costs. Also, family practice and internal medicine programs may not be comparable to psychiatry because of structural differences and because of psychiatry programs' larger and more elaborate curricula. Before the implementation of the new Medicare guidelines around teaching-physician billing, psychiatry residents may also have had less direct bedside teaching than one might find in family practice or internal medicine, electing instead to discuss patients in formal supervision at other times.
One can, however, gain a rough idea of general costs from a review of existing literature. Table 1 outlines the range of costs associated with family practice and internal medicine residency training programs. Family practice literature contains the most information about residency costs. All costs are presented as indicated in the articles and updated (parentheses) to reflect consumer price index increases since the index year. Barnett et al. (9) examined the costs involved in family practice residency training in eight California public hospitals for the 1984–85 fiscal year. Using Medicare cost data, they reported per- resident costs ranging from $54,865 base-year dollars ($83,354 in 1998 dollars) to $94,889 base-year dollars ($144,231 in 1998 dollars). These data did not reflect clinical revenue produced by residents. Analyzing data from the state of California that allocated costs differently, the same programs ranged from $46,837 ($66,508) to $84,936 ($129,102). This illustrates that the way in which costs are allocated can influence total costs.
Colwell and Glenn (10) reported 1978–79 costs for a family practice residency program at the University of Missouri–Columbia. With 30 residents, "total identifiable" costs were $1,541,000, including faculty and support staff salaries, overhead costs of two practice centers, travel, and supplies. Using these figures, the per-resident cost is $51,366 ($96,208). In 1995, Jones et al. (11) calculated the costs of an academic family practice residency located in a community health center (CHC). These appear to be only a portion of total training costs, as they reflect only the amount of time spent at the CHC. Nevertheless, in the CHC, PGY-I residents were calculated to incur costs of $20,162 ($21,280), PGY-II residents cost $39,555 ($41,255), and PGY-III residents, $51,598 ($53,816).
In a comprehensive investigation in 1994, the Massachusetts League of Community Mental Health Centers (12) used an accounting firm to demonstrate costs of family practice residency programs in rural and urban sites. Total costs were not calculated, but net costs associated with residency training (including resident-generated patient revenue) were $35,826 ($39,372) at the rural site and $23,675 ($26,018) at the urban site. Stern et al. (13), in 1975, calculated yearly costs of approximately $39,240 ($86,720) per resident in a series of settings affiliated with Harvard Medical School. Using 1992 data, Zweifler (14) calculated a much higher figure of $96,000 ($111,168) per resident at two university-affiliated family practice programs in California.
For internal medicine, only one report in the literature outlines training costs. Diamond et al. (15) reported on costs and revenues associated with expanding an internal medicine residency from 22 residents in 1988 to 48 residents in 1991. Although they were not explicitly reported, costs can be calculated from their data. With 22 residents, the cost per resident appears to be $85,208 ($114,349). With 48 residents, costs drop to $81,482 ($97,696). This appears reasonable, since with increasing numbers of residents, administrative and other costs are spread over a larger group, decreasing the cost per resident.
There is an impressive variation in program costs among studies. Some can be attributed to varying methodologies, resulting in differences in what is counted as a cost and in how various costs are calculated. Regional differences in the consumer price index may play a role as well. Also, some costs can be allocated to more than one cost center. The classic example of this problem is costing out clinical teaching time. Clinical work could be allocated to a clinical cost center, but if teaching medical students or residents is involved, some part of this time could be allocated to a teaching cost center. It is entirely unclear how much time one might allocate for teaching costs, since teaching and clinical care are often done simultaneously. Moreover, neither direct nor indirect GME payments to hospitals reflect a real measurement of this cost in any direct way. Indirect GME payments are intended to pay for the added costs (not teaching) of patient care in hospitals. It is entirely unclear what these costs are or how to account for them, and there is substantial controversy regarding whether payments adequately reimburse hospitals for these costs (16).
|
METHODS
|
The goals of this project were to:
- Determine the total costs of psychiatry residency training for one Midwestern adult residency program.
- Calculate the "hidden" costs of training, including direct as well as "in-kind" expenses involved in supervision and unpaid didactic instruction by volunteer faculty.
- Contrast these costs with those provided in the literature for training programs in other specialties.
Subjects
Subjects who participated in the study were either full-time faculty or unpaid community-based faculty termed clinical faculty in Michigan State University's Department of Psychiatry Residency Training Program. The survey was sent to 36 faculty members, 11 who were full-time and 25 based in the community who were not paid by the university. This included all faculty members who might provide bedside supervision, office supervision, or seminar didactics. MSU faculty who did not teach or work with residents were excluded from the survey.
The authors' residency program is community- based, with few full-time faculty involved in clinical teaching. Only one full-time faculty member directs an inpatient unit on which residents rotate. All other units are directed by clinical faculty members who derive income from sources other than the university. One other full-time faculty member directs the on- campus outpatient clinic on a one-half FTE (full-time equivalent) basis. Some clinical faculty members in the community do outpatient case supervision for residents. Other outpatient teaching is done in the local Community Mental Health Center (CMHC) by CMHC-employed psychiatrists who are clinical faculty members. Clinical faculty members who provide outpatient supervision do so by meeting with residents for 1 hour per week. Costs reflected for this supervision are based on the average reimbursement in the Department of Psychiatry for 1 hour of clinical time, since clinicians doing supervision would likely want to be reimbursed for their time at a rate similar to this.
Design
A survey instrument was designed to allow faculty to estimate time allocations associated with direct supervision and teaching. Subjects were asked to complete a blank weekly schedule partitioned into sections with beginning and ending times for the day. Headings to which time could be attributed included 1) office-based supervision time; 2) telephone time; 3) reading/preparation time; 4) clinical teaching rounds; and 5) other. Direct resident contact was defined as time spent face-to-face with the trainee, including bedside rounds, office supervision, and any informal conversations about educational issues. Other time spent on educational activities involving residents was defined to include telephone contact, reviewing resident charts or other records, and preparation for supervision or courses. Participants were requested to document time in quarter-hour blocks spent with residents each day for the period March 10–17, 1996. The survey was returned anonymously. A second wave was sent to nonrespondents after 1 month. Although any 1-week sample period may be atypical, the months of March and April do not contain any unusual events such as vacations, holidays, or activities that pull residents from rotations and might bias data collection.
The psychiatry residency program budget provided data on residency costs for paid faculty, as well as administrative, support staff, recruitment, and supplies. These include
- 0.5 FTE training director, associate training director, and clinic director salaries
- Full-time salary for residency education secretary, 0.5 FTE for a graduate assistant
- Other department salaries, including the chair, at 0.15 FTE, and business office personnel at 0.25 FTE
- Travel costs for one annual meeting for the training director, associate training director, and chief resident
- Supplies and services, including a) recruitment b) office supplies c) phones and pagers d) accreditation fees e) PRITE (Psychiatry Resident-In-Training) examinations f) other resident fees
- 0.2 FTE teaching time for didactic seminars
Other costs include malpractice insurance, resident stipends, and fringe benefits. Many didactic seminars are provided without charge by clinical faculty psychiatrists and other mental health professionals who teach 1 to 2 hours per year. A few individuals who teach for lengthy periods of time and who travel from Detroit are reimbursed for their seminar expenses, reflected in Item 6 above. Administrative support staff salaries, including fringe benefits, were based on the amount of time estimated to be devoted to residency education, as well as some teaching time that was reimbursed. Although for completeness the cost of clinical teaching rounds is included in Table 2, this cost is not included in the total cost calculation for the residency. HCFA (Health Care Financing Administration) regulations regarding faculty billing require faculty members to bill only for time spent with patients. Presumably, faculty time spent with patients would be billed whether or not the resident was present. In our setting, the two faculty members who do clinical teaching with residents each would continue to carry similar caseloads without residents, and therefore their billing and reimbursement are not affected by having residents with them. Other full-time and clinical faculty members responding to this survey have established caseloads that do not change on the basis of house-staff availability.
At the time of this report, the Department of Psychiatry did not pay overhead costs such as rent, utilities, custodial services, other maintenance costs, or distributed billing costs for the group practice. These costs were calculated based on per-square-foot costs for similar clinical practice space and would come to a total of $69,922.
The department operates a small outpatient psychiatry clinic exclusively for training purposes. This clinic operates on a sliding-scale fee system and covers the costs of front-desk staff and a few supplies. For the year, it brought in no net revenue but did not account for any significant losses, either. The costs of faculty teaching time are accounted for in the survey, rather than broken out for the clinic as another cost.
Data Analysis
A full yearly time commitment was based on a 46-week work year. In order to calculate yearly supervision costs, a figure of $67.00 was used as an average for an hour of supervision on the basis of the average reimbursement to the Department of Psychiatry. To derive yearly costs, the average number of supervisory and telephone-and-so-forth hours per week was multiplied by the average $67.00 the department would have to pay for the supervisor's time and then multiplied by a 46-week work year.
|
RESULTS
|
The overall response rate for the faculty questionnaire was 77% (N=28). The eight nonrespondents are a problem to account for. Given that we could not identify respondents, it is not possible to give a best estimate of their clinical time. However, on the basis of the pattern of returned surveys, we believe that the time not accounted for by these eight individuals could not be more than $20,000 to $40,000. However we are unable to give a confident estimate within this range. Table 2 demonstrates the average time reported by faculty members spent in residency-related activities. Supervision time includes individual supervision, group supervision, and various other clinically-related seminars not directly a part of the didactic seminar schedule. "Clinical teaching rounds" includes time spent on wards in traditional teaching rounds and some clinic teaching with residents, as well as consultation/liaison psychiatry supervision.
The numbers of respondents vary for each category because not all faculty members engage in all resident teaching activities. The largest time commitment is spent in clinical teaching rounds, often with one resident at a time. Individual supervision is also a large category and is required and extensive in our program. Given that this is non-reimbursed time, if faculty were to use this time in reimbursed patient care, the remuneration to the department would be significant. This is a cost the department pays to have a training program. The second-largest time commitment is actually that of reading/preparation time for residency teaching and supervision.
The psychiatry residency program budget can be seen in Table 3. Although this is a budget, it is in fact an accurate reflection of costs, since it is updated periodically by correcting expenditures in the process of tracking line-item expenses. Other true costs of the program are those we attempted to calculate on the basis of our time survey data (Table 2). These include supervisory time and clinical teaching time, reading and preparation time, telephone time, and an "other" category. Teaching time in seminars done by volunteer clinical faculty members is another "hidden" cost. At $67.00 per lecture hour, this cost is estimated at $23,048 for 344 hours of seminars.
The total identified program cost of supporting 20 residents, excluding only clinical teaching time for the reason cited earlier, is $1,563,193, or $78,159 per resident.
Table 2 and Table 3 reveal that the largest proportion of costs is that of personnel (personnel excluding residents, resident stipends and fringe benefits, office-based supervision, reading/preparation time, and other), accounting for 84% of costs. An additional significant expense for the 1996 year was malpractice insurance. Because of changes in the risk profile of the faculty group practice, these charges were larger for 1996 than in preceding years and are expected to decrease in subsequent years.
|
DISCUSSION
|
The derived program costs are consistent with those of many earlier studies in other specialties. They also approximate the costs derived by many hospitals from the direct GME reimbursement portion of Medicare. When the indirect portion, added for increased patient care costs involved in training residents (usually 1 to 2 times the direct costs) is added, it is clear that there are many hospitals that have made substantial margins from GME dollars. It is also probable that hospitals use enhanced margins from this revenue flow to help support money-losing but necessary services such as emergency departments.
Cost structures are going to vary by program, but there are some common features. Personnel are clearly going to represent the largest proportion of costs in any program. Some programs will have larger faculty and facility costs, and many will likely derive large proportions of their funding from inpatient activities. A part of these personnel costs are the significant amounts of time and money expended by faculty in various otherwise not well-accounted-for costs of psychiatry residency training. Faculty members report a large amount of time, and thus monetary commitment, in reading/preparation time and less in telephone time. Preparation time may include a diverse set of activities. More time is likely needed to develop and update seminars than for supervision. Faculty members may also be including time spent reading medical literature related to questions raised by residents. The total for this unreimbursed time spent in residency-related activities represents a large expenditure and resource that could potentially be used for patient care or other activities if there was no residency program. Where to allocate these resources is a value judgment to be made by departments.
With this kind of cost structure and the costs involved in training residents, vanishingly few graduate medical education programs could be supported without an external source of GME funding. The recent MedPAC report advocating the use of an "enhanced patient care adjustment" does not take into account faculty time and costs as are demonstrated in this report (17). Also, given the cost structure of programs, budget cuts are problematic. The only area in which cuts are possible in any substantial way is personnel. To accomplish this, programs reduce either faculty or resident positions. Faculty members produce revenue, but in many programs, residents provide patient care on inpatient units that would be difficult for faculty members to provide. Furthermore, departments may have to hire nurse-clinicians or physician assistants if resident numbers are reduced. Salary savings may be questionable. A number of hospitals have dropped out of the New York Demonstration project after concluding that replacing house staff with other health care professionals would actually incur increased costs (18). The rolling- average method for GME payments postpones decreases in GME dollars due to decreases in house- staff numbers, but eventually, GME dollars will decrease. Given this, it is not clear that cutting house- staff positions will result in cost savings. On the other hand, resident position cuts may make sense from a workforce point of view, since there are fewer applicants then positions, and inpatient psychiatry bed numbers have been decreasing so that fewer residents may be needed to staff these units.
Free-standing psychiatric hospitals and units in general hospitals that are DRG (diagnosis-related groups)-exempt represent another problem for programs and for some hospitals. In both these situations, institutions receive only direct GME funding. Training programs located in these institutions are at a disadvantage because of this funding difference. Community-based programs are also disadvantaged because of the structure of GME, since, generally, hospitals, and not programs, are eligible for GME funding. This structural issue is both discriminatory and limiting to types of rotations. It also probably inhibits the development of novel program designs in other than hospital settings. The new HCFA regulations addressing funding of a few consortial GME ventures do begin to address this issue.
A detailed examination of residency program costs should help departments make decisions regarding how to structure and continue to fund their training programs as well as negotiate funding with hospitals. The often-not-inconsequential contributions of volunteer faculty and other "in-kind" contributions can be understood and perhaps appropriately recognized. It is also useful to demonstrate to many institutions, not the least of which is the federal government, that quality training carries a significant price tag.
Limitations
The time portion of this study was conducted during only 1 week. Although this was a typical week of the program, a longer sample period is desirable. Also, the eight nonresponders must result in some factitious decrease in costs, although it does not appear that any were providing a great deal of supervision.
Given that the MSU program is community- based, it may not be representative of academic health center-based programs. These programs may have both larger revenues, for example, from inpatient units, and larger costs. Programs located in areas where the cost of living is greater than in the Midwest are likely to find that many of their costs are larger on a per-unit basis than those of this Midwestern program.
Finally, the thorny problem of how to "cost out" time that faculty spend with patients teaching residents is entirely unclear. It is not clear whether our inpatient-unit faculty member would be more or less productive without residents and thus whether resident teaching represents a cost in terms of clinical time for which we need to account.
|
ACKNOWLEDGMENTS
|
The authors thank Jason Dahn for statistical support and Carol York and Alice Nobis for secretarial assistance.
|
REFERENCES
|
- Magen JG, Banazak DA: Graduate medical education financing in psychiatry. Academic Psychiatry 1995; 19:6-11[Abstract]
- Iglehart J: Support for Academic Medical Centers. New Engl J Med 1999; 341:299-304[Free Full Text]
- AAMC Reporter 1995; 4(9):1-50
- Forman S, Kerr WB, Mullins CB, et al: Academic medical centers, proposed Medicare cuts, and the emerging price- competitive environment. Acad Med 1995; 70:510-511
- American Psychiatric Association: APA Memorandum 11/ 26/97
- Whitcomb ME, Cleverly WO: Financial performance of academic medical center hospitals. Acad Med 199; 68:729-731
- Flint L, Flint CB: Academic surgical practices at the dawn of health reform. Ann Surg 1994; 220:374-381[Medline]
- Association of American Medical Colleges: AAMC Data Book January 2000
- Barnett PG, Midtling JE, Burnett WH, et al: Cost-benefit analysis of California family-practice residencies. J Fam Pract 1989; 28:567-572[Medline]
- Colwill J, Glenn JK: Patient care, income, and the financing of residency education. J Fam Pract 1981; 13:529-536[Medline]
- Jones TF, Culpepper L, Shea C: Analysis of the cost of training residents in a community health center. Acad Med 1995; 70:523-531[Medline]
- Massachusetts League of Community Health Centers and Center for Community Education, Research, and Service: Analysis and Report on the Cost of Medical and Health Profession Training at Community Health Centers. Alexander, Aronson, and Finning, Certified Public Accountants, March 15,1994
- Stern RS, Jennings M, Delbanco T, et al: Graduate medical education in primary care. New Engl J Med 1977; 279:638- 643
- Zweifler J: Balancing service and education: linking community health centers and family-practice residency programs Fam Med 1993; 25:306-311
- Diamond HS, Fitzgerald LL, Day R: An analysis of the cost and revenue of an expanded medical residency. J Gen Intern Med 1993; 8:614-618[Medline]
- Cameron JM: The indirect costs of graduate medical education. New Engl J Med 1985; 312:1233-1238
- Rethinking Medicare's payment policies for graduate medical education and teaching hospitals. Medicare Payment Advisory Commission, August 1999
- Council on Graduate Medical Education: Physician Workforce Policies: COGME Recommendations for the Twenty- First Century: Implications of Recent Developments. Rockville, MD, Dept. of Health and Human Services, 1998
This article has been cited by other articles:
|
|
|
|
 
S. I. Kramer
Where There's Smoke, Can There be Fire?
Acad Psychiatry,
November 1, 2008;
32(6):
458 - 459.
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. J. Prochaska, S. C. Fromont, D. Leek, K. Suchanek Hudmon, A. K. Louie, M. H. Jacobs, and S. M. Hall
Evaluation of an Evidence-Based Tobacco Treatment Curriculum for Psychiatry Residency Training Programs
Acad Psychiatry,
November 1, 2008;
32(6):
484 - 492.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. Dunstone
A Pilot Study for Integrating Volunteers With Core Psychiatry Faculty
Acad Psychiatry,
August 1, 2005;
29(3):
262 - 266.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. A. Walter, P. B. Rosenquist, and G. Bawtinhimer
Distance Learning Technologies in the Training of Psychiatry Residents: A Critical Assessment
Acad Psychiatry,
March 1, 2004;
28(1):
60 - 65.
[Abstract]
[Full Text]
[PDF]
|
|
|
Get information about faster international access.
Privacy Policy
Copyright © 2000
Academic Psychiatry.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
