PEDIATRICS Vol. 108 No. 1 July 2001, pp. 79-84

Characteristics and Financial Performance of a Pediatric Faculty Inpatient Attending Service: A Resource-Based Relative Value Scale Analysis

Sanford M. Melzer, MD, Richard A. Molteni, MD, Edgar K. Marcuse, MD, MPH, and Frederick P. Rivara, MD, MPH

From the Department of Pediatrics, University of Washington School of Medicine and Children's Hospital and Regional Medical Center, Seattle, Washington.

	ABSTRACT

Top Abstract Methods Results Discussion References

Objectives.  In many children's hospitals, inpatient attending physician services are provided by academic faculty who function as part-time inpatient specialists or hospitalists. Although some have claimed that hospitalist care can reduce length of stay and total hospital resource use and expenses, there are few benchmarks or data regarding physician productivity or the characteristics and financial performance of these programs. The resource-based relative value scale (RBRVS) is a valuable tool for developing national benchmarks and comparing the financial performance of inpatient programs at varying daily census and reimbursement levels. The objectives of this study were to 1) describe physician productivity on an inpatient service as measured by total relative value units (TRVUs) and professional charges, 2) determine whether inpatient collections were adequate to support faculty salaries for the time spent attending, and 3) develop a model to evaluate financial performance of inpatient programs at varying census and TRVU reimbursement levels.

Methods.  A retrospective review of hospital discharge and faculty practice billing data between June 1997 and July 1998 was conducted in a general medical service in a regional, 208-bed, university-affiliated children's hospital in the Pacific Northwest.

Results.  Of 4113 patients who were admitted to the children's hospital general medical service during a 12-month period, faculty part-time hospitalists (N = 28) served as the attending physician for 1738 (42%). On an annual basis, faculty attended for an average of 29.1 days (median: 21.0; range: 7.0-97.0), with an average daily patient census (ADC) of 7.2 (median: 6.5; range: 2.8-12.0). Inpatient attendings billed for 1738 initial visits and 3957 subsequent visits. Total physician productivity for the inpatient attending group during 1 year included 12 085 TRVUs and gross professional charges of $777 743. The average payment, or conversion factor (CF), was $24.46/TRVU (71% of Medicare CF). The cash collection rate was 38%, reflecting a payor mix that included 54% Medicaid, 28% commercial payors, 12% health maintenance organization, and 6% other payors. On a weekly basis, physicians generated an average of 109 TRVUs and collected $2665 in cash. The average salary cost per RVU was $23.40, and weekly faculty salary and benefit expenses were $2550. After operating expenses and academic taxes totaling 24% were deducted ($5.87/TRVU), RBRVS-based payments and cash collections covered 79% of average faculty weekly salaries. Financial modeling showed that either an average CF of $31/TRVU or an ADC of 9 patients per day on the inpatient service would be required to generate sufficient revenue to support physician salaries and operating expenses.

Conclusions.  For a faculty inpatient attending service in a children's hospital with an ADC of 7, a $24.46 RBRVS-based CF payment is inadequate to support faculty salaries and operating expenses for the time spent attending. Inpatient services in similar payor environments with comparable expenses and staffed by faculty who care for fewer than 9 patients per day will not cover typical faculty salary costs and operating expenses.  Key words:  inpatient care, resource-based relative value scale, hospitalist.

Historically, in nonrural hospitals, children have been cared for by community pediatricians or by academic pediatric faculty who serve as inpatient physicians. Recently, declining pediatric inpatient hospitalization rates, decreasing experience of community physicians with hospital care,¹ capitation contracts, and increasing productivity demands have led pediatricians to develop alternative models for inpatient care. In response to physician concerns and competitive pressures, many children's and community hospitals have considered programs that provide inpatient attending services by a designated cadre of physicians.^2-4 Physicians who staff these programs may be community-based "hospitalists," commonly defined as physicians who spend at least 25% of their time based in a hospital setting, or academic faculty who serve as part-time hospitalists to provide inpatient care on a rotating basis.

Proponents argue that hospitalists enhance quality and efficiency because care is provided by physicians who are experienced in the care of hospitalized children.² Hospitalist programs also may control resource use and decrease costs of care and length of stay when compared with traditional attending models,⁵ in part because of enhanced availability and communication with referring physicians.⁶ In markets where inpatient services that are paid at a diagnosis-related group (DRG)-based per case rate or paid per diem represent a significant portion of payments, such expense control takes on increased importance.

Hospitalist programs also are promoted as a mechanism for hospital marketing and to increase market share. In competitive health care environments, some hospitals may view pediatric hospitalists as a way to attract more pediatric patients by supporting local pediatricians and surgeons who refer to the hospital but choose not to care for their own inpatients and by providing a broader scope of pediatric services on the inpatient unit, emergency department, delivery room, and nursery.

Although there is a growing literature describing the clinical, organizational, and financial aspects of inpatient programs that serve adult patients,^7-9 there is a paucity of information regarding pediatric hospitalists, inpatient attending programs, and the physician productivity and payment rates needed to support the salaries of hospitalist physicians. Nonetheless, some have raised concerns that these programs are not a cost-effective way to provide 24-hour care, especially without concurrent resident physician coverage.¹⁰

Development of inpatient productivity and financial benchmarks has been hampered by the lack of a standardized financial reporting format. Presentation of program financial performance and physician productivity in terms of gross charges, "usual, customary, and reasonable" fees, or local collection rates is insufficient to allow programs in different payor environments to compare their financial experience. The Health Care Financing Administration (HCFA) resource-based relative value scale (RBRVS) represents a uniform national benchmark for physician payment and as such is a valuable tool that facilitates such comparison.¹¹ This fee schedule, developed initially for reimbursement in the Medicare program but now in increasing use in pediatrics,¹² provides an objective measure of physician work and may be used to describe and compare physician productivity in different clinical settings. The RBRVS system also is used to standardize reimbursement by assigning relative value units (RVUs) to each current procedural terminology (CPT) code. The most commonly used RVU measure is "total RVUs" (TRVUs), which take into account the physician work, practice expenses, and professional liability that are associated with specific CPT codes. The TRVU in turn is adjusted further for specific regional variations in health care costs through the use of a geographic practice cost index. Under the RBRVS system, payment is determined by applying a conversion factor (CF)a value that converts the TRVU into payment amounts for reimbursing physicians.

The Medicare CF used nationally is set annually by the Congress as part of the Medicare budget process. Among other payors that use the RVU system, wide variation exists in CFs that are used to establish payments. This state-by-state variation is particularly striking within the Medicaid program. Conversion of payments by health plans into RVUs not only provides a method to evaluate and negotiate contracts but also is an effective tool to describe and contrast payments in health care environments that are characterized by differences in payor mix and reimbursement rates.

In this study, we describe the characteristics and financial outcomes of a pediatric inpatient attending service that is staffed by faculty who serve as part-time hospitalists. Specifically, we report on the physician productivity and financial performance of the inpatient service using both the RBRVS system and more traditional measures, including patient visits, professional charges, and collection rates. We determine whether inpatient collections were adequate to support faculty salaries for the time spent attending and present a financial model to explore the volume and payor conditions under which revenues that are generated by inpatient attending services would be sufficient to support faculty salaries for time spent attending.

	METHODS

Top Abstract Methods Results Discussion References

Inpatient Practice Characteristics

We analyzed the general medical service of a regional, 208-bed, university-affiliated, freestanding children's hospital in Seattle, Washington, with more than 11 000 annual discharges. The inpatient service was divided during the year into 2 or 3 teams, each with a single attending. Approximately 95% of the attending rotations were staffed by faculty attendings; community physicians served as inpatient attendings during the remaining rotations. Faculty inpatient workload was calculated from a review of inpatient attending schedules for July 1997 to June 1998. CPT codes, charges, and collections were obtained from faculty practice billing records. Faculty salaries and benefits were obtained from the Department of Pediatrics annual budget.

Calculation of TRVUs and CF

TRVUs were determined for each of the initial and subsequent hospital care CPT codes used by the inpatient attending group, according to 1999 standard HCFA methodology.^13,14 The TRVU value is derived by combining an assignment of a standard, HCFA-determined numeric value for physician work, practice expenses, and malpractice expenses for each CPT code. The TRVU values then were adjusted for each component of the TRVU with the use of the HCFA geographic practice cost index. The CF per TRVU for the inpatient service was calculated by dividing the total cash collections by TRVUs.

Financial Performance of the Inpatient Service

Financial performance of the faculty inpatient service at varying daily census was calculated with the use of standard profit/loss accounting methods. In the cash flow model, revenues were projected with the use of calculated ratios of revenues per patient at different volumes, and expenses were based on actual salary, benefits, and operating expense rates. In the RVU-based model, revenues were projected with the use of calculated values for daily TRVU/patient at specific CF payment levels.

	RESULTS

Top Abstract Methods Results Discussion References

Of 4113 pediatric patients who were admitted to the children's hospital general medical service during the 1-year study period, faculty part-time hospitalists (N = 28) served as inpatient attendings for 1738 patients (42%). The faculty hospitalist group included 13 general pediatricians and 15 pediatric specialists: neurodevelopmental (5), infectious disease (4), rheumatology/immunology (4), genetics (1), and nephrology (1). Fourteen (50%) were associate professors, 7 were professors, 5 assistant professors, and 2 were instructors.

Practice Characteristics of the Inpatient Attendings

Faculty members served as inpatient physicians on the inpatient teams for a total of 814 days during the year-long study period, with a mean of 29.1 days per physician on service (median: 21.0; range: 7.0-97.0). The average daily census (ADC) on the faculty service was 7.2 patients per attending per day (Table 1). Inpatient physicians billed for 5695 inpatient visits, including 1738 (31%) initial visits using CPT codes 99221 to 99223 and 3957 (69%) subsequent visits using CPT codes 99231 to 99233 (Table 2). Among initial and subsequent visit codes (Fig 1), most were of intermediate complexity (61% and 66%, respectively), whereas 34% of initial visit codes and 9% of subsequent visits were of high complexity.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Distribution of CPT codes for initial and subsequent hospital care.

Physician Productivity, Revenues, and Expenses

Total physician productivity for the inpatient attending group during the 1-year period included 12 085 TRVUs and gross professional charges of $777 743. On a weekly basis, physicians generated an average of 109 TRVUs and $7014 in gross billings. The overall collection rate on professional charges was 38%, reflecting deductions from revenue, including contractual allowances on government contracts, negotiated discounts from payors, bad debt, and free or partial-pay care, leaving only $295 595 for salary and benefit distribution. The payor mix included 54% Medicaid, 28% commercial, 12% health maintenance organization, and 6% other payors. Average weekly net collections were $2665 per physician, and the average CF payment was $24.46/TRVU (71% of Medicare). For 48 weeks per year (1.0 full-time equivalent [FTE]) of inpatient attending, the net revenue (before operating expenses) was $127 920. After operating expenses, medical school dean, and departmental taxes totaling 24% ($5.87/TRVU) of collections were deducted, net revenue was $18.58/TRVU. Weekly faculty salary and benefit expenses were $2550, and average salary cost per RVU was $23.40. Weekly cash collections thus covered 79% of average faculty weekly salaries.

Using the typical time spent for each CPT code as described in the CPT manual, we calculated the total and daily physician work effort devoted to inpatient care (Table 2). The 12 085 RVUs generated 3163.3 hours, or 3.9 hours of billable work per day, during the 814 days of faculty inpatient attending time. Physicians generated 3.8 RVUs per hour, and with the use of our actual conversion factor of $24.46, this resulted in a reimbursement rate of $93/hour of billable time. With the assumption that a single 40-hour FTE employed physician works 44 weeks or 1760 hours, a total of 1.79 FTEs dedicated to inpatient care would be required to cover the clinical (nonteaching) physician work on this inpatient service.

Financial Performance of the Faculty Inpatient Practice

Using a cash flow model including actual salaries and daily revenues, we calculated the projected profit/loss of the hospitalist service at varying ADC volumes (Table 3). At an ADC of 7 patients (the actual ADC during this study), the weekly deficit was $525, or $27 300 annually. This deficit decreased to $235 per week with an average inpatient census of 8. An ADC of 9 patients per day on the inpatient service provided sufficient revenues to support the average physician weekly salaries and operating expensesthat is, it became revenue neutral.

The financial performance of the program at different ADCs and CFs was modeled. For salary costs of $2550/wk ($365/d) and operating costs of 24% of cash collections, a conversion factor of $31 would be required at an ADC of 7 to provide sufficient revenue to cover faculty costs. Alternatively, assuming a CF of $25, an ADC of 9 patients per day on the inpatient service would be required to generate sufficient revenue to support physician salaries and operating expenses. These results were extrapolated to provide a simple calculation of collections per week at varying ADC and RVU conversion rates (Table 4). For example, an inpatient program with physician salary costs and operating expenses of $2800/wk would generate sufficient revenue to cover its costs with an ADC of 5 and a CF of $36. However, in a less favorable payor environment with an average CF of only $28, an ADC of 7 patients would be required to achieve the same financial result.

	DISCUSSION

Top Abstract Methods Results Discussion References

Using an RBRVS-based analysis, we showed that for this faculty inpatient attending service in a children's hospital with an ADC of 7, a CF payment of $24.46 per TRVU for professional services is inadequate to support actual faculty salaries, benefits, academic taxes, and operating expenses. Inpatient revenues covered only 79% of the physician salary and benefit costs, with a weekly deficit of $525 per physician. Our financial modeling indicates that for a program with a similar expense structure, a conversion factor of $31/TRVU or an ADC of 9 would be required to generate sufficient revenue to support physician salaries and operating expenses.

Pediatric physicians in both academic and nonacademic settings have been faced with a declining reimbursement rate over the past few years. Our finding that the inpatient attending service was operating at a financial deficit was not unexpected, especially given our high proportion of Medicaid patients. The net financial results of a pediatric inpatient service reflect important and consistent influences on revenue and expense. In certain managed care environments, professional revenues for specific inpatient services may be bundled (along with hospital charges) into a combined "per-case" rate. More typical, inpatient revenues are based on a discounted fee-for-service fee schedule. Many plans have converted to an RVU-based fee schedule with payment set as a percentage of Medicare reimbursement. In this study, our CF of $24.46 is substantially less than the $34.24 national HCFA Medicare CF in effect at the time of this study. Our payment rate falls between the $44/RVU CFs offered by some Seattle commercial payors and the $22.00/TRVU that was being paid by Medicaid in Washington state at the time of this study.¹⁵ During the study period, this Medicaid reimbursement rate was insufficient to cover basic salary and benefit costs ($23.20/TRVU) and failed to allow for academic taxes and other operating expenses, such as malpractice insurance or billing and collection services. The finding that Medicaid payments for inpatient services do not cover physician salaries is concerning and once again emphasizes the need for equitable reimbursement rates for children's services to ensure access and quality of care for inpatient services for this underserved child population.

The use of an RBRVS-based analysis highlights the importance of diligent and accurate coding practices, to ensure that physicians receive adequate credit for the work (including time, effort, stress, and training) that they provide. Among our physicians, CPT codes reflecting the highest complexity of medical decision (99223 and 99233) were used for 34% of initial and 9% of subsequent hospital care encounters. This coding pattern differs somewhat from that described in a national survey of pediatric physicians, who reported using level 3 hospital codes in 36% of initial hospital care encounters and 32% of subsequent hospital care encounters.¹⁶ The underuse of higher-complexity CPT codes (compared with national norms) for subsequent visits among our physicians points out the need for ongoing physician education in appropriate coding techniques. The goal of these activities is a complex one; to maximize charge capture by ensuring that the coding reflects the physician effort accurately and also to ensure compliance with the specific and complex documentation requirements of payors and federal agencies.

Because the CPT codes used by our faculty have assigned typical times, the RBRVS system also permits an analysis of workload based on an estimate of actual time spent performing physician work. Although time alone may not be used to determine level of coding, it can be used to estimate the commitment (excluding teaching time) to provide these services that our faculty reported in a single year. This method indicates that our faculty spent 3163.3 hours generating 12 085 RVUs, or 3.8 RVUs per hour. Understanding physician production in terms of RVU/hour facilitates planning and managing an inpatient service. Once the salary per physician is established and the RVU production per hour is established, the CF necessary to cover physician expenses can be determined. The actual CF then can be calculated with the use of a blended payor mix and reimbursement rate.

In inpatient programs, expenses that are not covered by the prevailing CF would require hospital subsidy. The degree to which a given program requires subsidy is dependent on program costs, payor mix, blended CF, patient volumes, productivity, and charge capture. Higher patient volumes, more favorable physician contracts, increased physician productivity, and improved charge capture each would decrease the need for hospital subsidy, and any revenue in excess of physician salary and expenses may be used as salary incentives for employed physicians.

Inpatient physician productivity also can be measured in terms of patient census. There are few benchmarks for our faculty hospitalist inpatient workload. In a survey of 365 practicing hospitalists on adult medical services, Lurie et al¹⁷ found that the average workload was 13 inpatients. Pediatric programs that are seeking to enhance the financial performance of their inpatient service may wish to structure their inpatient services to increase individual attending daily workloads, thus increasing the likelihood that an individual physician will generate sufficient revenue to cover his or her salary costs.

We believe that although many of these findings can be generalized to other academic programs, certain key determinants of financial performance will differ between our academic program and other care models. For example, our physicians did not routinely bill for certain procedures that generally were performed by residents, such as lumbar punctures or bladder taps, or use available codes for discharge day services. In contrast to the faculty physicians in our practice who provide only inpatient care, community-based inpatient physicians or hospitalists may generate additional revenue through coverage of the emergency department, delivery room, or normal nursery or through provision of consultation services to family physicians and surgeons. Although additional revenues may be available to community-based hospitalists, inpatient physician salary costs also may be higher outside of academic programs. In a survey of 372 hospitalist physicians, most of whom were in community practice, Lindenauer et al¹⁸ reported a mean income of $143 000. However, only 5% of the respondents in that study were pediatricians, and the reported average salary may not reflect the experience of the typical pediatric hospitalist. The $108 700 we report as our mean faculty salary is within 10% of the national median for associate professor-level general pediatric faculty members in pediatric departments.¹⁹ Another important determinant of salary cost in a hospitalist program is the requirement for 24-hour coverage. In the academic model, residents often provide this coverage, whereas in community hospitals, 24-hour coverage is provided by in-house hospitalists, sometimes in combination with nurse practitioners or physician assistants. Therefore, a week of continuous coverage may require up to 4 individuals in a community hospital, whereas this same coverage can be provided by a single daytime attending combined with in-house residents in a teaching setting. When our hospital codes were converted to typical times, we found the need for 1.8 FTE physicians to cover the clinical (nonteaching) needs of these patients, with residents providing 24-hour coverage. Finally, operating expenses in academic inpatient practice may be higher as a result of the support of academic departments through an allocation of a portion of professional revenues.

We showed that this inpatient attending service in a children's hospital operates at a deficit of $26 000 per year of coverage per physician. This estimate assumes that all activity and funding sources during clinical rotations derive from clinical work. We are aware that some faculty have other sources of salary support that continue during "on-service" time, and in these cases, our methods may lead to an overstatement of the program losses. However, we also found that practice expenses often do not cover total practice costs. This finding, combined with our experience with funding of clinical faculty members, leads us to believe that this study actually underestimates the complete program costs in academic centers. Few individuals are able or willing to serve exclusively as inpatient hospitalists for 12 months per year, and for these faculty members, time spent teaching and performing other academic and administrative functions often is undercompensated.

What are the implications of these data for community or children's hospitals that are seeking to evaluate the financial performance of a pediatric hospitalist service? First, programs should model their revenue carefully with the use of ADC, hourly and daily RBRVS-based physician productivity (from all clinical sources), and specific RVU CFs and collection rates that reflect their unique payor mix. Children's hospitals with a majority of Medicaid patients should anticipate that the financial advantages of high patient volumes might be offset significantly by low RVU-based CFs and/or highly discounted fee-for-service contracts. After modeling revenue, cost must be considered carefully. In academic programs, these considerations should include operating expenses such as billing, collection, insurance, academic and practice plan taxes, and benefits, as well as any inherent inefficiencies in the academic staffing model. Community hospital programs need to evaluate salary costs carefully, taking into consideration whether in-hospital 24-hour attending care is required and whether revenues derived from providing overnight coverage will offset large additional salary expenses.

After analyzing projected physician productivity and reimbursement, we believe that in many cases, hospitals will find that physician productivity, patient volumes, and associated revenues are insufficient to cover professional costs of a hospitalist program. Under this circumstance, is hospital support of these programs justified? In academic programs, many would argue that the training and service mission of inpatient faculty attendings might justify the additional cost. If so, how will costs be apportioned among departments, divisions, and the hospital? Children's hospitals and community hospitals also may justify the expense of these programs, especially those that are seeking to increase market share, by encouraging pediatric referrals by local physicians and surgeons.

Although we provided a method to evaluate the financial outcomes of hospitalist programs using tools such as the RBRVS, decisions to support these programs should go beyond financial performance and ideally must be based on the medical outcomes of this model of care. Future research to evaluate clinical outcomes, health status, patient satisfaction, and resource use^20,21 is critical as we seek to train, deploy, and finance hospitalists (or academic faculty serving as hospitalists) in pediatric inpatient settings.

	ACKNOWLEDGMENT

We thank Alex Dunne for providing invaluable assistance in preparing this article.

	FOOTNOTES

Received for publication Aug 2, 2000; accepted Nov 11, 2000.

Address correspondence to Sanford M. Melzer, MD, Children's Hospital and Regional Medical Center, Medical Administration, Mail Stop CH-41, 4800 Sand Point Way, N.E., Seattle, WA 98105-0371.

	ABBREVIATIONS

DRG, diagnosis-related group; HCFA, Health Care Financing Administration; RBRVS, resource-based relative value scale; RVU, relative value unit; CPT, current procedural terminology; TRVU, total relative value unit; CF, conversion factor; ADC, average daily census; FTE, full-time equivalent.

	REFERENCES

Top Abstract Methods Results Discussion References

1. Melzer SM, Grossman DG, Rivara FP Physician experience with inpatient care in Washington state. Pediatrics 1996; 97:65-70 [Abstract/Free Full Text]
2. Bellet PS, Wachter RM The hospitalist movement and its implications for the care of hospitalized children. Pediatrics 1999; 103:473-477 [Free Full Text]
3. Wachter RM Hospitalists: their role in the American health care system . J Med Pract Manage 1996; 13:123-126
4. Heldrich FJ When pediatric hospitalists make sense. Contemp Pediatr 1999; 16:117-137
5. Halpert AP, Pearson SD, LeWine HE, Mckean SCW The impact of an inpatient physician program on quality, utilization, and satisfaction. Am J Manag Care 2000; 6:549-555 [Medline]
6. Bellet PS, Whitaker RC Evaluation of a pediatric hospitalist service: impact on length of stay and hospital charges. Pediatrics 2000; 105:478-484 [Abstract/Free Full Text]
7. Diamond HS, Goldberg E, Janosky JE The effect of full time faculty hospitalists on the efficiency of care in a teaching hospital. Ann Intern Med 1998; 129:197-203 [Abstract/Free Full Text]
8. Freese RB The Park Nicollet experience in establishing a hospitalist system . Ann Intern Med 1999; 130:350-354 [Abstract/Free Full Text]
9. Meltzer D, Morrison A, Guth T, Effects of hospitalist physicians on an academic general medicine service: results of a randomized trial. J Gen Intern Med 1999; 14:112 [CrossRef][Medline]
10. Roberts KB, Rappo PD A hospitalist movement? Where to? Pediatrics 1999; 3:499
11. Medicare RBRVS. The Physician's Guide. Chicago, IL: American Medical Association; 1998:1-125
12. Resource Based Relative Value Scale Project Advisory Committee Issues in the application of the resource based relative value scale system to pediatrics: a subject review. American Academy of Pediatrics. Pediatrics 1998; 102:996-998 [Abstract/Free Full Text]
13. Medicare RBRVS. The Physician's Guide. Chicago, IL: American Medical Association; 1999
14. 1999 St. Anthony's Complete RBRVS. Reston, VA: St. Anthony's Publishing, Inc; 1999
15. Medicaid Reimbursement Survey. Chicago, IL: American Academy of Pediatrics; 1999
16. Agrawal B, Tomany S. Pediatric Service Utilization, Fees, and Managed Care Arrangements. 1999 Report Based on 1997 Data. Chicago, IL: American Academy of Pediatrics; 1999
17. Lurie JD, Miller DP, Lindenauer PK, The potential size of the hospitalist workforce in the United States. Am J Med 1999; 106:441-445 [CrossRef][Medline]
18. Lindenauer PK, Pantilat SZ, Katz PP, Wachter RM Hospitalists and the practice of inpatient medicine: results of a survey of the National Association of Inpatient Physicians. Ann Intern Med 1999; 130:343-349 [Abstract/Free Full Text]
19. Medical School Pediatric Faculty Compensation Survey1999. Association of Administrators in Academic Pediatrics; 1998
20. Showstack J, Katz P, Weber E Evaluating the impact of hospitalists. Ann Intern Med 1999; 130:376-381 [Abstract/Free Full Text]
21. Michota FA Evidence based evaluation of the hospitalist model of care. J Clin Outcomes Manage 2000; 7:49-53