A Model of Determining a Fair Market Value for Teaching Residents: Who Profits?
,
* Department of Anesthesiology and Critical Care Medicine
Nemours Office of Operational Assessment, Nemours Children’s Clinic, Alfred I. duPont Hospital for Children, Wilmington, Delaware
Thomas Jefferson University School of Medicine, Philadelphia, Pennsylvania
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Context. Centers for Medicare & Medicaid Services (CMS) Health Resources and Services Administration Children’s Hospitals Graduate Medical Education (GME) Payment Program now supports freestanding children’s teaching hospitals.
Objective. To analyze the fair market value impact of GME payment on resident teaching efforts in our pediatric intensive care unit (PICU).
Design. Cost-accounting model, developed from a 1-year retrospective, descriptive, single-institution, longitudinal study, applied to physician teachers, residents, and CMS.
Setting. Sixteen-bed PICU in a freestanding, university-affiliated children’s teaching hospital.
Participants. Pediatric critical care physicians, second-year residents.
Main Outcome Measures. Cost of physician opportunity time; CMS investment return; the teaching physicians’ investment return; residents’ investment return; service balance between CMS and teaching service investment margins; economic balance points; fair market value.
Results. GME payments to our hospital increased 4.8-fold from $577 886 to $2 772 606 during a 1-year period. Critical care physicians’ teaching opportunity cost rose from $250 097 to $262 215 to provide 1523 educational hours (6853 relative value units). Residents’ net financial value for service provided to the PICU rose from $245 964 to $317 299. There is an uneven return on investment in resident education for CMS, critical care physicians, and residents. Economic balance points are achievable for the present educational efforts of the CMS, critical care physicians, and residents if the present direct medical education payment increases from 29.38% to 36%.
Conclusions. The current CMS Health Resources and Services Administration Children’s Hospitals GME Payment Program produces uneven investment returns for CMS, critical care physicians, and residents. We propose a cost-accounting model, based on perceived production capability measured in relative value units and available GME funds, that would allow a clinical service to balance and obtain a fair market value for the resident education efforts of CMS, physician teachers, and residents.
Key Words: economic balance point • fair market value • opportunity cost • relative value unit • return on investment for resident education
Abbreviations: CMS, Centers for Medicare & Medicaid Services • DME, direct medical education • GME, graduate medical education • HRSA, Health Resources and Services Administration • IME, indirect medical education • PGY-2, postgraduate second year • PICU, pediatric intensive care unit • ROEI, critical care physicians’ investment return on teaching residents in the PICU for the service that residents provide in the PICU • RORI, CMS investment return on GME reimbursement for PICU resident education that occurs in our freestanding children’s teaching hospital • RROS, resident investment return for service provided in our PICU for both clinical teaching by critical care physicians and salary from DME funds • RVU, relative value unit
Medicare and Medicaid training grants for primary care medicine and dentistry, as well as cross subsidies from patient-care revenue, fund graduate medical education (GME) in the United States.1 Centers for Medicare & Medicaid Services (CMS) GME payments to individual teaching hospitals consist of payments for direct medical education (DME) and for indirect medical education (IME). The DME payments are designed to fund residents’ salaries and benefits, teaching physicians’ compensation, and teaching materials. The IME payments are calculated to include a portion of the indirect costs caused by the hospital’s patient-care operating inefficiencies associated with training residents.2 Indirect expenses stem from relatively intense use of diagnostic services by residents, decreased productivity of nurses and other staff who help teach residents, and use of expensive medical technology for research and educational purposes.3 Teaching hospitals are not accountable for how they use IME funds.3 Before 2000, freestanding children’s teaching hospitals received yearly an average $374 per resident in GME funds from Medicare, whereas nonchildren’s teaching hospitals received an average $87 034 per resident.4 This disparity in the level of federal funding for freestanding pediatric versus nonpediatric teaching hospitals is attributable to few Medicare patients being cared for in children’s hospitals.
In June 2000, the Health Resources and Services Administration (HRSA) announced the Children’s Hospitals GME Payment Program.4 The HRSA Children’s Hospitals GME Payment Program authorized $280 million for fiscal year 2000 and $285 million for fiscal year 2001 divided as one third for direct and two thirds for indirect expenses to be given to eligible children’s hospitals that operate GME programs.4 The HRSA determines the specific allocation that individual freestanding children’s teaching hospitals will receive. The fiscal year 2000 national average GME payment to freestanding children’s hospital teaching programs was determined to be $71 709 per pediatric resident.5
In January 1992, the Health Care Financing Administration replaced a physician payment system that was based on customary, prevailing, and reasonable charges with a Resource-Based Relative Value Scale physician fee scale derived from the relative value of services provided. The relative value unit (RVU) of each physician service is derived from the amount of work that goes into the service, the practice expense associated with the service, and the professional liability expense for providing the service. The RVU is multiplied by a geographic adjustment factor for each Medicare fee-schedule area and translated into a dollar amount by an annually adjusted conversion factor. This dollar amount is the reimbursement a physician receives for the RVU.6 On the basis of previous experience7 and recent billing records, we assign an average of 4.5 RVU to a critical care physician’s hour of clinical service. The time in hours that pediatric critical care physicians could potentially spend providing clinical care but instead spend teaching residents in the pediatric intensive care unit (PICU) is the opportunity time. The cost of opportunity time that pediatric critical care physicians spend teaching residents in the PICU is the number of teaching hours provided to residents in the PICU multiplied by both 4.5 RVU per hour and the prevailing Medicare conversion factor.
Residents provide service in the PICU such as gathering patient histories and performing physical examinations, completing discharge paperwork, entering orders, assisting with procedures, and providing ongoing care of patients under the direct supervision of the pediatric critical care physician. The quantification of this resident service was previously reported in a 1993 Urban Institute report8 wherein the 2055 RVU clinical value for work a resident provides yearly in a teaching hospital was described as the 1992 resident GME of $63 712 divided by the 1992 Medicare conversion factor of $31.00. The current PICU resident service worth in RVU can be derived from the hospital GME revenue directed toward the PICU residents and the prevalent payment per RVU.
Fair market value is the price at which an asset or service passes from a willing seller to a willing buyer. It is assumed that both buyer and seller are rational and have a reasonable knowledge of relevant facts.9 The CMS sells GME funds to teaching hospitals and buys an educational product from critical care physician teachers that is directed at PICU residents-in-training. Teaching hospitals buy IME funds from the CMS and sell an environment to residents in which they can obtain a pediatric critical care educational experience. Critical care physicians buy PICU resident service, which is bolstered by teaching hospital DME funds, and sell their clinical expertise to the CMS-funded PICU residents-in-training. Residents buy a PICU educational experience with teaching hospital DME and IME funds and sell various services to the critical care physicians during their PICU clinical rotation.
In this study, we analyze the impact that the CMS HSRA Children’s Hospitals GME Payment Program has on resident teaching efforts in our PICU. By applying simple cost-accounting methodologies to the critical care teaching service, CMS, and resident cost drivers, we propose a model that achieves a fair market value balance among the CMS investment return for their GME in PICU resident education, the critical care physicians’ investment return for teaching residents in the PICU, and the residents’ investment return for service in the PICU. We suggest that our model can be adapted to other teaching services.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The actual HRSA Children’s Hospitals GME Payment Program GME, DME, and IME allocations for fiscal years 2000 and 2001 to our freestanding children’s teaching hospital are documented.10,11
Educational activity provided by critical care attending physicians to postgraduate second-year (PGY-2) residents at our 16-bed, university-affiliated, freestanding children’s teaching hospital’s PICU was prospectively tracked from July 1, 1999 to June 30, 2000. Educational activity was credited in the following manner. One hour per day was credited for morning clinical rounds. One hour each day was credited for bedside supervision by the attending critical care physician. One hour each night was credited for bedside supervision by the attending critical care physician who was available by phone and present in the PICU for all admissions and for changes in patients’ clinical course. The following sessions were given by separate, designated teaching physicians and credited for 1 hour each: PICU introduction session for new resident groups, mock practice resuscitation teaching scenarios, and interactive teaching sessions. The time spent administering the PICU resident rotation and maintaining our PICU resident educational database was also prospectively tracked. Resident hours in the PICU were calculated from on-call schedules and individual program ongoing commitments. Teaching material costs were considered negligible for this study.
The Appendix contains definitions and calculations that can be used to clarify the concepts of the CMS return on GME investment in PICU resident education, the teaching physicians’ return on resident teaching investment, and the residents’ return on service in exchange for teaching and salary. The Appendix also contains definitions and calculations used in a cost-accounting model of the interactions among the CMS, critical care physicians, and residents when providing clinical education to residents in a PICU.
A basic and fully adaptable cost-accounting model that includes variables relevant to physician teachers, residents, and CMS was created using Microsoft Excel (Microsoft Corporation, Bellevue, WA). Input variables included the following: resident demographics; PICU allocation of hospital GME, DME, and IME reimbursement; number of residents per PICU rotation; allocation of residents’ salary and benefits to the PICU; PICU educational hours; RVU value assigned to a critical care clinical hour; Medicare dollars per RVU; and an assumed practice premium contract payment per RVU. Model variables can be adjusted as needed. Outcome measures included the following: critical care physician teaching opportunity time and associated opportunity cost; calculated teaching salary for pediatric critical care physicians’ efforts; the value of the residents’ service that they provide during their PICU rotation; net financial value that residents bring to the PICU during their rotation; the CMS investment return for their GME reimbursement in PICU resident education (RORI), the critical care physicians’ investment return for teaching residents (ROEI) in the PICU as a function of the service residents provide in the PICU, and the residents’ investment return for service (RROS) provided in the PICU as a function of the teaching opportunity time of the critical care physicians and resident salary; the margin (profit or loss) for CMS and the PICU teaching service as well as the balance (service balance in absolute dollars) between these margins. For baseline analysis, we assumed that reimbursement did not have a premium over Medicare. The RORI minus the ROEI was calculated.
A positive difference between RORI and ROEI implies that CMS "receives a better return" for their dollar, whereas a negative difference implies that the pediatric critical care teaching service "receives a better return" for its dollar. Service balance is the absolute dollar difference between the CMS teaching investment margin and PICU teaching service investment margin. If the service balance is positive, it favors CMS; if the service balance is negative, it favors the teaching service. A higher RROS more favorably benefits residents.
Sensitivity analyses (the impact of input changes on output) were performed using different GME reimbursements, DME percentages of GME, teaching hours, number of residents taught, and service balance. Using the solver function of Microsoft Excel, we allowed the RVU per hour, the teaching hours per year of a service, or the percentage of GME that was allocated to DME or all 3 to change while keeping the other variables within the model constrained. We defined and determined economic balance points, which are the points at which the RORI, ROEI, and RROS are equal within particular scenarios.
Data were analyzed using 
2 test, analysis of variance, or paired t test as appropriate and were reported as means ± standard deviation. A value of P < .05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
From July 1, 1999 to June 30, 2000, 6 critical care physicians (4.4 full-time employee level) documented 1523 (1297 teaching and 226 administrative) hours equivalent to 6853 RVU that were devoted to educational activity for 49 PGY-2 residents (18 pediatric and 31 nonpediatric; Table 1). The same basic educational program framework continued through the academic year 2001–2002. Each resident spends an average of 261 (58% day, 42% night) hours in the PICU during his or her month-long rotation.
|
The fiscal year 2000 GME and DME revenues allocated for the PICU from the $557 886 CMS HRSA Children’s Hospitals GME Payment Program to our hospital were $15 208 and $4468, respectively (Table 2). The resident salary and benefit expense that was incurred by the PICU was $56 250. The total annual opportunity cost for teaching PICU residents was $250 907. The RORI was 528%. ROEI in the PICU in exchange for the service provided by residents was 76%. The RORI minus the ROEI and the $296 892 service balance imply that the CMS "receives a better return" for their investment than does the teaching service. Residents provided 6596 RVU and a net financial value of $245 964, which represented an RROS of 125%.
|
The fiscal year 2001 GME and DME revenues allocated for the PICU from the $2 772 606 CMS HRSA Children’s Hospitals GME Payment Program to our hospital were $77 017 and $22 628, respectively (Table 2). The resident salary expense that was incurred by the PICU was $56 250. In 2001, because of an increase in Medicare payment per RVU from $36.61 to $38.26, the teaching opportunity cost for critical care physicians and the residents’ net financial value increased to $262 215 and $317 299, respectively. The RORI dropped to 70%. ROEI increased to 100%. A –30% RORI minus ROEI slightly favored the PICU teaching service when 1523 teaching hours were provided per year (4 hours a day; Fig 1). The $186 434 absolute dollar difference between the CMS and PICU teaching service investment margins still favored the CMS at this level of resident education effort (4 hours per day; Fig 2). The RROS dropped to 100%.
|
|
Sensitivity analyses (Table 2, scenarios 1, 2, and 3) illustrate the impact of potential increased support for a freestanding children’s teaching hospital on the RORI, ROEI, and RROS. If a freestanding children’s teaching hospital could incrementally receive 1) the same resident GME payment that is now allocated to nonpediatric teaching hospitals, 2) the full GME funds for residents from nonpediatric teaching hospitals who are participating in a rotation at a children’s hospital, and 3) a DME payment that is 70.62% of the GME payment, then the physicians’ investment return would increase to 169%, but the CMS and residents’ investment returns would decrease to 46% and 70%, respectively. Sensitivity analysis illustrates that if a hospital provided 2580 teaching hours yearly (7.8 hours daily) to residents under the conditions of Scenario 3, then RORI, ROEI, and RROS may balance (Fig 3). At this level of teaching hours, the service balance would still favor the CMS (Fig 4).
|
|
Using the solver function of Microsoft Excel, we could not find a solution to balance RORI, ROEI, and RROS at the present fiscal year 2001 CMS HRSA Children’s Hospitals GME Payment of $2 772 606 to our freestanding children’s teaching hospital. However, a solution is possible if the percentage of GME payments allocated to DME payments is changed from 29.38% to 36%. This solution represents the economic balance points, which are the points at which the return on investment for CMS, physicians, and residents is equal for various degrees of service teaching hours and usual productivity per hour of the service (Fig 5). This suggests that it is possible to achieve a fair market value for the efforts of CMS, physician teachers, and residents in resident education.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Cost should be a factor but not the driving force in the process to develop high-quality resident education.12 However, because cost accounting and responsibility for teaching are becoming fiscal issues, flexible budget analyses must be formalized to determine the specific impact of teaching efforts.
The sense of privilege and the personal rewards of teaching will remain important, but uncompensated teaching time is likely to become an issue in an era in which financial compensation for physicians’ activities is more closely monitored.13
Postgraduate education for pediatricians is grounded in the perception of the specific health care needs of children and youth.14 Although critical care represents a small proportion of the total health care needs of children and youth, there is significant value for continued exposure of PGY-2 trainees to critical care during their residency programs.15 The difficulties involved in providing adequate funding for pediatric training programs have been addressed recently by The Future of Pediatric Education II Financing of Pediatric Education Workgroup16 and the American Academy of Pediatrics Committee on Pediatric Workforce.1 The present HRSA Children’s Hospitals GME Payment Program creates a new funding source for postgraduate pediatric education. Our study shows that the fiscal year 2001 $22 628 DME allocation to our PICU does not cover even the resident salary expenses, let alone the critical care physicians’ salaries for teaching. Teaching potentially could be a revenue-generating experience if our freestanding children’s teaching hospital received the same GME funds that residents at nonpediatric teaching hospitals receive, if GME funds were also available to our freestanding children’s teaching hospital from programs that send nonpediatric residents to our PICU for training, and if the percentage of DME per GME payments was increased to the present IME level of 70.62% per GME payment (Table 2, scenario 3).
The critical care hour reimbursement for teaching could also include a premium in addition to the Medicare rate. Premium denotes the percentage increase over Medicare RVU reimbursement that a local employer may add to the present Medicare RVU value. The premium would be an important factor in determining opportunity cost. However, in our model, a premium would also increase the value of residents’ service that they provide during their PICU rotations, which, in turn, would affect the critical care physicians’ investment return on teaching residents in the PICU for the service residents provide in the PICU.
Decreased teaching or the use of other teaching models may be acceptable alternatives to balancing the financial impact on our PICU resident rotation if educational outcomes and resident satisfaction are significantly better. But this linear relationship has yet to be validated.
Other PICUs may offer their own "homegrown" RVU system in which content and educational time commitment differ from our reported efforts. However, it is feasible that all PICU resident educational encounters can be converted to RVU equivalents and that pediatric critical care physician opportunity time for teaching PICU resident trainees can be measured and benchmarked as RVU. Two studies have shown how an RVU-like system can be used to compensate faculty involved with teaching critical care.17,18 We appreciate that it is difficult to separate simultaneous clinical and educational activities. We also appreciate that we may have underestimated the actual hours related to resident teaching activities. Reporting and conversion of hours and, thus, RVU for bedside supervision and clinical rounds could be misleading, but they are frequently used as leverage in budgetary and regulatory evaluations.
It is potentially misleading to assume that if a critical care physician is not teaching that he or she would be generating clinical revenue. Opportunity cost assumes that time spent teaching could be used instead for clinical revenue-producing activity. Whether the opportunity actually exists for additional clinical activity versus teaching is very dependent on the specific make up of units. However, physicians are always under pressure to participate in other clinical revenue-producing areas. The Leapfrog Group for Patient Safety19 initiative may increase the opportunity for opportunity cost because it has identified on-site intensive care unit physician staffing as a crucial, quality initiative.
As part of their educational experience, residents provide services to the PICU. There is an ongoing tension to balance residents’ clinical service and their educational experience in the PICU. We did not include an assessment of the indirect service provided by residents, which challenges critical care teachers to keep up-to-date with critical care issues.
Several other assumptions were not included in the construction of the model. Would the absence of residents at PICU patient rounds lead to more opportunity time for the critical care physicians to pursue additional revenue-producing clinical activities? Can GME funds allow a specific children’s hospital to recruit visiting faculty who could provide up-to-date educational material to both the faculty and residents in the intensive care unit?
We are challenged to maximize the individual resident’s critical care knowledge base and comfort level with critically ill children during the average of 261 hours (58% day, 42% night) that he or she spends in the PICU during our month-long rotation. There is potential to provide additional self-learning educational material through access to computer-assisted and Internet teaching modules. But these substitutes lack the interactive experience with the pediatric critical care physician that our previous residents valued highly in their educational experience.20
Because of the increasing financial burden of decreased reimbursements on physician medical practices, the pressure is mounting to decrease teaching time. Classically, teaching time is perceived as an under-reimbursed and undervalued service. This argument was correct for freestanding children’s teaching hospitals until GME reimbursement distribution was altered to include these hospitals. Now the argument focuses on how these GME funds could be more fairly distributed between physician teachers and teaching hospitals.
Using the current GME payment program information, we could not find an economic balance point at which the RORI, ROEI, and the RROS are equal. However, our model does demonstrate that if the percentage of GME as DME would increase from 29.38% to 36% at our present $2 772 606 GME hospital reimbursement, then economic balance points are achievable at which the investment return for physicians, CMS, and residents is equal for various degrees of service teaching hours and usual productivity per hour (Fig 5).
We offer a model that can be generalized to other resident teaching situations to help hospitals tailor their educational activity on the basis of available GME funds. The model allows one to change variables and then determine how many teaching hours would be recommended for the various realities of actual RVU per hour production to maintain the economic balance points. Our model allows a clinical service or department to set teaching time goals (not quality) that are based on the perceived production capability as measured in terms of RVU.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Our analysis demonstrates that the current CMS HSRA Children’s Hospitals GME Payment to our freestanding children’s teaching hospital produces an uneven investment return for the CMS, critical care physicians, and residents. Future studies are needed to verify whether our cost-accounting model can be generalized to other resident teaching situations to help set teaching time goals that are based on production capability as measured in terms of RVU and available GME funds.
|
|
ACKNOWLEDGMENTS |
|---|
Copies of the study model are available from the corresponding author on request.
We thank Dr Pamela Arn for reviewing the manuscript.
Dr Nadkarni is affiliated with the Department of Anesthesia and Critical Care, Children’s Hospital of Philadelphia, Pennsylvania. Dr Corddry is affiliated with the Anne Arundel Medical Center, Annapolis, Maryland.
|
FOOTNOTES |
|---|
Received for publication Jul 8, 2002; Accepted .
Reprint requests to (E.J.C.) Department of Anesthesiology and Critical Care, Alfred I. duPont Hospital for Children, Box 269, Wilmington, DE 19899. E-mail: ecullen{at}nemours.org
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
- American Academy of Pediatrics, Committee on Pediatric Workforce. Financing graduate medical education to meet pediatric workforce needs.
Pediatrics.2001; 107
:785
–789
[Abstract/Free Full Text] - Wray JL, Sadowski SM. Defining teaching hospitals’ GME strategy in response to new financial and market challenges. Acad Med.1998; 73 :370 –379[Web of Science][Medline]
- Nicholson S. The effects of Medicare payment subsidies to teaching hospitals. Issue Brief, Leonard Davis Institute of Health Economics. December 2001/January 2002;7
- National Archives and Records Administration. Federal Register.Monday, June 19, 2000, Vol. 65, no. 118, 37985 [00–15340]
- National Archives and Records Administration. Federal Register.Thursday, March 1, 2001, Vol. 66, no. 41, 12947 [01–5008]
- Thompson L, Risley T. Committee on Coding and Reimbursement. 2000 RBRVS, What Is It and How Does It Affect Pediatrics? Elk Grove Village, IL: American Academy of Pediatrics; 2000
- Cullen EJ, Lawless ST, Nadkarni VM, et al. Evaluation of a pediatric intensive care residency curriculum. Crit Care Med.1997; 25 :1898 –1903[CrossRef][Web of Science][Medline]
- Miller M, Welch W. Analysis of Hospital Medical Staff Volume Performance Standards: Technical Report. Washington, DC: The Urban Institute Working Paper; 1993:6201–6210
- Friedman JP. Dictionary of Business Terms. Hauppauge, NY: Barron’s Educational Series, Inc; 1994:216
- National Association of Children’s Hospitals. HRSA List of DME and IDME Payments. Alexandria, VA: National Association of Children’s Hospitals; 2000
- HRSA NEWS. HHS awards nearly $227 million to children’s teaching hospitals to ensure quality health care for America’s children. Available at: http://www.hrsa.gov/Newsroom/releases/2001%20Releases/childhospgrantsl.htm. Accessed March 21, 2002
- Yeh TS. Give a little, take a little: resident education in the evolving healthcare environment. Crit Care Med.1997; 25 :1782[CrossRef][Web of Science][Medline]
- Shea S, Nickerson KG, Tenenbaum J, et al. Compensation to a department of medicine and its faculty members for the teaching of medical students and house staff.
N Engl J Med.1996; 334
:162
–167
[Abstract/Free Full Text] - Behrman RE. Postgraduate education for pediatricians.
Pediatrics.2001; 107
:553
–557
[Free Full Text] - Spevetz A, Brilli RJ. General guidelines for resident teaching in critical care medicine. New Horiz.1998; 6 :255 –259
- Jones MD Jr, Boat T, Adler R, et al. Final Report of the FOPE II Financing of Pediatric Education Workgroup. Pediatrics.2000; 106(suppl) :1256 –1270
- Powner DJ, Rogers PL, Kellum JA. Compensation for teaching in critical care. Crit Care Med.2000; 28 :1612 –1615[CrossRef][Web of Science][Medline]
- Nutter DO, Bond JS, Coller BS, et al. Measuring faculty effort and contributions in medical education. Acad Med.2000; 75 :199 –207[Medline]
- The Leapfrog Group for Patient Safety. Available at: http://www.leapfroggroup.org. Accessed March 25, 2002
- Cullen EJ, Lawless ST, Nadkarni VM, McCloskey JJ, Corddry D. PICU resident learner (RL), resident educator (RE) and utility scores [abstract]. Pediatr Res.1999; 45 :76A
- National Archives and Records Administration. Federal Register.Tuesday, November 2, 1999, Vol. 64, no. 211, 59380–59590 [99–28367]
PEDIATRICS (ISSN 1098-4275). ©2003 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
