Practice Expense Costs in Pulmonary and Critical Care Practices
Is Providing Patient Care Still Economically Feasible?

	INTRODUCTION

TOP INTRODUCTION A SURVEY OF EXPENSES... DO MEDICARE REIMBURSEMENTS... CONCLUSIONS REFERENCES

Medicare reimbursements for clinical services serve as a foundation to examine payments to pulmonary and critical care physicians in the United States. Medicare beneficiaries comprise a large percentage of the patient population in the pulmonary and critical care specialty. In addition to setting the Medicare fee schedule for physician services, the Medicare program also influences the physician payment rates of commercial insurers. Many commercial insurance carrier fee schedules reflect a percentage (above or below) of the Medicare fee schedule (1). Because of its pivotal role in setting both Medicare and commercial insurer physician reimbursement rates, it is important to verify whether Medicare reimbursements cover the entire cost of providing physician services to Medicare beneficiaries.

Medicare Part B reimbursements for professional physician services are based upon three components: physician work, malpractice insurance, and practice expenses. The practice expense component includes nonphysician salaries and benefits, rent, depreciation, utilities, and all other expenses (except malpractice insurance) necessary to run the practice. On average, for any given professional service covered by Part B of the Medicare program, physician work accounts for 55%, practice expenses account for 42%, and malpractice insurance accounts for 3.2% of the total reimbursement (2).

In 1994 Congress enacted legislation directing the Health Care Finance Administration (HCFA) to develop a resource-based system for calculating and reimbursing the practice expense component of physician reimbursement by 1997 (3). Later legislation delayed the implementation date of the resource-based practice expense system to 1999 (4).

Initial efforts by HCFA to determine resource-based practice expense focused upon quantifying the exact costs for each specific physician service. This technique of directly quantifying costs for each physician service has been described as a "bottom-up approach." However, numerous problems plagued the results from the "bottom-up" approach, including biased inputs from the physician community, wide variation in the estimated costs of an individual procedure between two (or more) different specialties, failure to identify all the component costs of providing an individual service, and improper valuations attached to each individual component of a provided service. Because of these and other problems with the "bottom-up" approach, the General Accounting Office (GAO) recommended in 1998 that HCFA refine and improve the methodology for determining resource-based practice expense (5).

In 1998, HCFA proposed an alternative "top-down approach" to the development of resource-based practice expense (6, 7). The "top-down" approach utilizes a survey of practice expense costs collected in the annual American Medical Association socioeconomic survey of physicians, and redistributes practice expense costs using a specialty-allocation method. The "top-down" allocation method results in assignment of practice expense to each Medicare service by a Current Procedural Terminology (CPT) code within a specialty that bears little relation to actual cost. The "top-down" method fixes the total funding of practice expense costs to each individual specialty. Practice expense data provided by the physician community indicates that different specialties often have different practice expense costs for providing the same clinical service, defined by a Current Procedural Terminology code, under Medicare. Fixing the total funding of practice expense to each specialty prevents Medicare from paying different practice expense rates to different specialties that perform the same clinical service.

HCFA solicited advice and practice expense data from the physician community to develop this resource-based practice expense system. To respond intelligently to the request from HCFA for information, the American Thoracic Society 1998- 1999 Clinical Practice Committee recommended that a study be commissioned to collect practice expense data for pulmonary and critical care physicians. Performed by the Gary Siegel Organization, a complete copy of this Practice Expense Survey is available on the ATS web site (8).

The original study goal was to collect sample data on practice expense costs from the pulmonary and critical care community. These data were to help guide the ATS response to the resource-based practice expense method proposed by HCFA. However, the data collected are also useful for accurately measuring practice expense costs for pulmonary and critical care physicians. Measuring practice expense costs allows pulmonary and critical care physicians to (1) evaluate the margin of profit (or loss) from Medicare, and other Medicare fee schedule-based reimbursements; (2) accurately compare practice expense costs across the pulmonary and critical care community; (3) develop "best business practices" that reduce practice expense costs and thereby maximize profit potential; and (4) demonstrate that Medicare reimbursement for practice expense costs bear little relation to the actual practice expense costs incurred providing a clinical service.

	A SURVEY OF EXPENSES IN PULMONARY PRACTICES

TOP INTRODUCTION A SURVEY OF EXPENSES... DO MEDICARE REIMBURSEMENTS... CONCLUSIONS REFERENCES

The ATS practice expense study utilized teams of accounting professors who visited nine pulmonary practices, conducting a 12-wk on-site cost study. The teams interviewed physicians and office staff, observed the provision of clinical services in both the office and hospital setting, and reviewed financial records. The characteristics of these nine practices from three states (California, Illinois, and Maryland) are shown in Table 1. These nine practices reflect the diversity within the pulmonary and critical care specialty, varying in size and organizational structure, providing a different mix of clinical services, and practicing in different economic environments.

Activity-based cost accounting (9) was used to collect practice expense data in the nine practices. Developed in the early 1980s, activity-based cost accounting is a generally accepted and reliable accounting method that determines product and service costs more accurately than do traditional costing systems. Working with the physicians and their office staff, the accounting teams identified the unique processes employed in providing clinical care to patients. The identified processes included providing patient care in both office or hospital-based encounters, performing procedures or diagnostic tests, obtaining insurance authorization, billing and collecting, maintaining medical records, and practice management. The accounting teams then utilized tax and financial records to determine the total amount spent by each practice in the study on practice expense costs. Total practice expense costs were then attributed first to each of the office processes, and then to the categories of clinical services delivered. Individual clinical services, including visits and procedures in both the office and hospital settings, are described using the specific Current Procedural Terminology (CPT) codes (10).

The results provide an average cost for four classes of clinical services: office visits (Table 2), hospital visits (Table 3), hospital procedures (Table 4), and office procedures (Table 5). The costs for all types and levels within each class of service were averaged for several reasons. Trivially small differences exist between different levels of service in the costs of individual processes (e.g., 7 min for insurance authorization of diagnostic bronchoscopy [CPT code 31622] versus 8 min for similar authorization of bronchoscopy with transbronchial biopsies [CPT code 31628]). Although physician time and medical complexity remain vastly different between different levels of service (e.g., straightforward office visit [CPT code 99212] versus complex office consultation [CPT code 99245]), staff time differences remain relatively small (minutes, not hours) and overhead (the consumption of practice expense inputs) is virtually identical. For example, providing a 10-min or a 45-min office visit requires roughly the same time to greet the patient, obtain the medical record, prepare the examination room, escort the patient to the examination room, answer the patient's questions, clean the examination room following the patient encounter, prepare and submit the bill, and schedule subsequent appointments and testing. The data in Tables 2, 3, 4, and 5 suggest that small variations in practice expenses between different levels of service are outweighed by large variations in cost among different practice types and settings.

Most physicians in this sample see patients in both the office and hospital settings, and perform procedures in the hospital settings. Office-based procedures (e.g., nasopharyngoscopy and thoracentesis) were far less common among survey participants. Blank rows in Tables 3, 4, and 5 exist for Practice 4, which did not offer any hospital-based services.

	DO MEDICARE REIMBURSEMENTS COVER COSTS?

TOP INTRODUCTION A SURVEY OF EXPENSES... DO MEDICARE REIMBURSEMENTS... CONCLUSIONS REFERENCES

This determination of actual costs allows meaningful comparison with Medicare reimbursements. Table 6 compares actual costs incurred providing selected Medicare services with the Medicare reimbursement for those services. Examples are shown for a moderate complexity office visit (CPT code 99214), low complexity office visit (CPT code 99213), subsequent ventilator management (CPT code 94657), moderate complexity hospital visit (CPT code 99232), diagnostic bronchoscopy (CPT code 31622), and thoracentesis (CPT code 32000). The first column (subheaded Low) compares the Medicare reimbursement for the designated service with the actual practice expense cost incurred by the private practice with the lowest amount of expense. The second column provides a similar analysis for the private practice with the highest practice expense, and the third and fourth columns show the comparison with the mean practice expense incurred by private practices and for university-/hospital-based practices, respectively. The excess of Medicare reimbursement over actual practice expense cost for each service in any setting must cover both the malpractice component and the physician work component of the service.

Malpractice expenses of the study participants were estimated by the following method: the malpractice relative value unit (RVU) (1) was divided by the total RVU for an individual service; this proportion was then multiplied by the Medicare reimbursement for the procedure or office visit. This estimated malpractice expense ensures that the time costs of malpractice insurance equal present Medicare charge-based reimbursements for malpractice. This assumption is certainly not true as HCFA changed from charge-based to resource-based malpractice expense RVUs beginning in 2000 (3). Nonetheless, this assumption allows us to crudely estimate malpractice expenses associated with a specific CPT code (10) and assess whether reimbursements cover costs for Medicare beneficiaries.

The moderate complexity office visit (CPT code 99214) is an office-based service provided by virtually every pulmonary physician who cares for outpatients. Estimating malpractice insurance at $1 leaves an adequate profit margin for pulmonary physicians to care for Medicare beneficiaries in the outpatient setting. However, recalculating these values using the $44 reimbursement for a low complexity office visit (CPT code 99213) would leave the average university-based practice and many high-expense private practices operating at a loss.

For subsequent ventilator management (CPT code 94657), one of the most common hospital-based services provided by pulmonary and critical care physicians, total Medicare reimbursement is $54. This reimbursement is quite similar to the total Medicare reimbursement ($58) for CPT code 99232, a moderate complexity hospital visit, with similar associated practice expense costs. For subsequent ventilator management, the mean actual private practice expense is $32, leaving $22 to cover malpractice insurance and physician compensation. Using the method described above, the estimated malpractice insurance expense is again $1, with $21 remaining to compensate a provider. If this hospital-based visit required 60 min of provider time, the physician would earn $21/h. For university-based practices, the practice expense costs incurred by providing CPT code 99232 exceeds the Medicare fee, indicating a net loss rendering this service.

In contrast, performing a hospital-based procedure such as bronchoscopy (CPT code 31622) appears to be appropriately compensating physicians for the work involved. Total reimbursement for this code is $236. Even in the university- or hospital-based practice, estimating $10 for malpractice insurance, $167 would remain for physician compensation. Nonetheless, bronchoscopies account for a very small proportion of the total clinical services provided by pulmonary and critical care physicians.

Of the practices surveyed, relatively few provided office-based procedures. Nonetheless, the small quantity of available data from the survey suggests that performing thoracentesis (CPT code 32000) in the office setting may be practical only for low-cost private practices. The survey results clearly demonstrate the practice expense costs for thoracentesis varied widely among the pulmonary practices participating in our survey. Such widely disparate costs may be expected among pulmonary practices reflecting their efficiency in performing a thoracentesis, their yearly volume of thoracenteses, their relative capacity costs (staff and facility), and their patient volume as reflected by the size of their referral network and the degree of competition in their region.

	CONCLUSIONS

TOP INTRODUCTION A SURVEY OF EXPENSES... DO MEDICARE REIMBURSEMENTS... CONCLUSIONS REFERENCES

The survey results show that for selected CPT codes (e.g., 32000), total Medicare reimbursement may not fully reimburse physicians for the practice expense of providing the clinical service. For these services, physicians are providing care to Medicare beneficiaries at a loss. For other services (e.g., CPT codes 31622 or 99214), Medicare reimbursement allows physicians to provide the service to Medicare beneficiaries at a profit. Other services (e.g., CPT codes 94657 or 99232) are profitable only in practices that have comparatively low practice expense costs. However, in all three scenarios Medicare reimbursement for practice expense costs has little relation to the actual amount of practice expense costs incurred to provide these services.

Wide variations exist among the costs within the various practices examined in the survey. In general, university-based practices always had the highest associated costs. Possible explanations might include the inefficiencies of clinical practice incurred from attention to the research and teaching missions of academia. However, our survey examined a very small number of practices that may not truly reflect the representative costs of academic practice. Nonetheless, in an era when many academic health systems remain under severe financial threat, the choice between fiscal solvency and the academic missions are likely to have an impact on patient care (11, 12).

Wide differences also appeared in the related practice expenses between hospital-based (e.g., bronchoscopy) and office-based (e.g., thoracentesis) procedures. Much of this cost variation likely reflects whether a facility (hospital) or the practice owns and maintains the equipment required for the procedure. However, as resource-based practice expenses are fully implemented over the next 2 yr (6, 7), the physician professional reimbursement for facility-based procedures will dramatically fall, reflecting the equipment costs being borne by the facility in this circumstance. These differential payments for the same clinical service, depending upon the site of the service (facility- versus office-based), are likely to induce changes in the delivery of care to Medicare beneficiaries. For example, bronchoscopy may be performed in an office setting far more commonly in the future, as the additional reimbursement garnered in the office may adequately offset the additional practice and malpractice expenses.

The survey results have implications for the access of Medicare beneficiaries to clinical care. Geographic variations among Medicare payments may not adequately compensate physicians for regional variation in the overhead costs of providing services (13). Such inadequate compensation may be most problematic in rural areas, where a low frequency of certain procedures would magnify the problems of fixed overhead costs, leading physicians to not offer certain services and thereby limit the access of Medicare beneficiaries to clinical care. However, our data do not directly address such geographic differences in either practice expense costs or Medicare reimbursements.

RVU-based reimbursement formulas based upon global specialty practice expenses ("top-down" methodology) currently being implemented by HCFA (6, 7) do not reflect the actual consumption of practice resources. These arbitrary formulas potentially distort practice expense reimbursement amounts to ultimately bear little relationship to the actual amounts expended via practice. This possibility raises serious questions about the propriety of these formulaic measures for setting practice expense reimbursement rates.

Using the detailed model available from the web site (8), any practice can determine its unit costs to provide clinical care. With appropriate random sampling of sufficient numbers and diversity of pulmonary and critical care practices, it is possible to create a database that can ultimately determine the average practice expense cost for providing each service in the United States. This database could help pulmonary and critical care physicians assess which Medicare services can be provided at a profit. Furthermore, such a database will enable physicians to compare their practice expenses with those of their colleagues. By comparing average costs among practices, physicians and practice managers can develop methods to reduce practice expenses and therefore maximize profit.

However, because Medicare reimbursement for practice expense is not based on actual expenses with use of the "top-down" methodology (6, 7), Medicare practice expense reimbursement by service (CPT code) will vary. Nonetheless, authentic practice expense will be similar for two closely related services (e.g., CPT codes 31623 versus 31624). The present estimates of practice expense emphasize to providers and practice managers alike the importance of cost management. With Medicare and other insurer payments fixed or falling, running a profitable practice will require knowledge and control of practice expenses.

	Footnotes

Correspondence and requests for reprints should be addressed to Scott Manaker, MD, PhD, 100 Centrex Building, Department of Medicine, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, PA 19104-4283. E-mail: scott.manaker{at}uphs.upenn.edu

	References

TOP INTRODUCTION A SURVEY OF EXPENSES... DO MEDICARE REIMBURSEMENTS... CONCLUSIONS REFERENCES

1. Medicare RBRVS:The physician's guide. Chicago, IL: American Medical Association; 1999.

2. Report to Congress: Medicare Payment Policy, Medicare Payment Advisory Commission, March 1999, p. 118.

3. Social Security Act Amendment. Public Law 103-432, Section 121, 1994.

4. Balanced Budget Act. Section 4505, 1997.

5. Report of the GAO to the Committees on Commerce, Ways, and Means, and Finance, March 1998.

6. Health Care Finance Administration. Notice of Proposed Rule Making, CFR 30818, 1998.

7. Health Care Finance Administration. Final Rule, 63 CFR 58814, 1998.

8. //http:www.thoracic.org

9. Holrngren CT, Sundem GL, Stratton WO. Introduction to management accounting, 10th ed. Englewood Cliffs, NJ: Prentice-Hall; 1996, 134- 145, 502-510, 533-540.

10. CPT 1998: Physician's current procedural terminology. Chicago, IL: American Medical Association; 1998.

11. Eisenberg L. Whatever happened to the faculty on the way to the Agora? Arch Intern Med 1999; 159: 2251-2256 [Free Full Text].

12. Iglehart JK. Support for academic medical centers: revisiting the 1997 Balanced Budget Act. N Engl J Med 1999; 341: 299-304 [Free Full Text].

13. Lewis JD, Asch DA. Barriers to office-based screening sigmoidoscopy: does reimbursement cover costs? Ann Intern Med 1999; 130: 525-530 [Abstract/Free Full Text].