Is employing physicians a smart strategy or a financial bubble?

U.S. hospitals and health systems report substantial operating losses from the medical practices they own or operate. National surveys indicate that health system-owned practices lose hundreds of thousands of dollars each year per employed physician.1 The losses, amplified by the large number of physicians in these systems, indicate that health systems are losing tens of millions of dollars operating their physician enterprises.

As inpatient margins continue to shrink, and health systems report the need to cut staff and “tighten their belts” to offset system-wide losses, we need to ask if employing physicians is a smart strategy or a financial bubble that will eventually burst and bring financial ruin to entire health systems. Examining the nation’s largest repository of medical group financial data helps to better understand the true magnitude and causes of the reported losses and to judge their potential impact on the parent organizations. The survey data also allow an assessment of whether the price of employing physicians fits the classic definition of a financial bubble, in which the cost of acquistion significantly exceeds the asset’s intrinsic value.

Physicians employed by hospitals and health systems lose money

In recent years, hospitals and health systems in the United States have expanded their acquisition of physician practices and employment of physicians. The latest AMA Physician Practice Benchmark Survey reported that in 2018 the percentage of physicians who were employed exceeded the percentage of physician owners and that the distribution of physicians shifted to large practices and hospital-owned practices.2 The reasons hospitals and health systems acquire practices and employ physicians vary considerably. Some hospitals and hospital systems describe their rationale for employing physicians as stemming from a desire to improve patient care through clinical integration, while others appear to have less noble strategies of preempting competition and directing employed doctors to utilize the hospital’s or system’s frequently underutilized imaging, laboratory and inpatient services. Additionally, health systems receive a strong economic benefit from inpatient admissions and services from their hospitals as documented in the 2019 Merritt Hawkins Physician Inpatient/Outpatient Revenue Survey, which reported an average annual net revenue of $2,378,727 generated by physicians across all specialties for their affiliated hospitals; some specialties produce more than $3 million for their hospitals.3 Securing the fealty of the hospital staff by employing physicians is an important aspect in understanding the economics of hospital systems.

Even if it is not explicitly stated in their strategic plans, many health systems recognize that their owned practices provide the opportunity to increase total organizational revenue through collecting technical and professional fees for physician services. Recent evidence indicates that such financial integration has led to higher total cost of care and increased total healthcare spending.4,5

The first wave of practice acquisition occurred in the early 1990s when hospitals and their systems initiated a strategy of purchasing primary care practices to compete for capitation insurance contracts.6 Almost immediately, most systems discovered that hospital-owned practices were not as efficient or as profitable as independent, physician-owned groups, and that practices became less efficient and less profitable after hospitals acquired them. The Advisory Board, which in 1995 promoted a strategy of purchasing physician practices and being paid on capitation,7 four years later published Stopping the Bleed, Reversing Losses on Owned Practices,8 reversing the advice by suggesting that hospitals divest their physicians and return to fee-for-service (FFS) payment.

Other key studies include:

• In 2012, Jeff Goldsmith reported that hospitals’ economic loss per physician averaged $212,000.”9
• A 2014 survey of Kentucky hospitals found that 58% of hospitals in the state reported annual losses of more than $100,000 per physician in their owned practices — an increase of 17% over the prior year.10
• A Congressional Research Service 2013 review of prior studies noted that hospitals lose $150,000 to $250,000 per year for the first three years they employ a doctor.11 

 
Current publications continue to report substantial losses. As Timothy Smith wrote in 2019, median loss exceeds $350,000 per full-time-equivalent (FTE) physician in hospital-owned surgical single specialty practices.12

Understanding physician production

A popular belief persists that losses in hospital-owned practices are the result of poor management and lower productivity by hospital-employed physicians. People subscribing to this belief allege that hospitals do not know how to run physician practices and employed doctors lack incentives to work as hard as their private practice peers. These beliefs are reflected in articles published 15 years ago, describing the “theory of the business” of a physician practice and showing that it is substantially different from that of a hospital.13,14 These articles described how most physician practices, especially in primary care, operated on a much smaller scale than hospitals, almost at a retail level with a high number of relatively short patient encounters with low-dollar charges, whereas hospitals’ policies are designed to optimize a system of fewer, higher-intensity patients and much higher charges. 

The belief that employed physicians are less productive than physicians who own their practice contends that employed physicians lack the entrepreneurial spirit that drives them to see more patients, expand services and improve efficiency. This perspective likely stems from the design of many physician employment contracts in the first wave of practice acquisition in the 1990s, which had minimal performance requirements or financial incentives to be productive. In 2002, Douglas Conrad and colleagues found that in that era “hospital ownership of the group is negatively related to RBRVS-measured productivity.”15 However, more current information suggests that this has changed as hospital systems moved away from straight salary for their doctors in favor of incentive-based compensation methodologies. Evidence of the changes in hospital-employed physician productivity published in 2019 described how primary care physicians in hospital-owned practices report more work RVUs (wRVUs) than their peers in physician-owned medical groups.16

Consistent with the shifts in productivity and management practice, our examination of data from a large cross section of physician- and hospital-owned practices show that most of the losses attributed to the hospital-employed physicians are accounting for differences in revenue and expense reporting and do not necessarily reflect substantial reductions in systemwide profits.

Study methodology

For more than 50 years, MGMA has collected and reported an extensive set of metrics describing total financial performance of medical groups of all types and specialties in the MGMA Cost and Revenue Survey.17 Data published in 2016 and 2018 Cost and Revenue Survey Reports are displayed in the tables examining the financial performance of independent multispecialty practices and those that are part of a health system. In addition to the published metrics, we used ordinary least squares multiple regression of the 2016 survey database to explore the productivity and financial performance of multispecialty practices owned by physicians compared to practices owned by hospitals and health systems using IBM® SPSS® Statistics Version 25.0.

We believe that our analysis is the first to look at the issue by comparing the financial information submitted by a large cross section of physician-owned and hospital-owned practices and examining sources of revenue, business office performance, and other comprehensive financial information.

Examining the data

Table 1 provides a side-by-side comparison of the financial performance of physician-owned and hospital-owned multispecialty medical practices reported in the MGMA 2016 and 2018 Cost and Revenue surveys (reflecting financial performance for CY/FY 2015 and 2017). From these reported numbers, it is apparent that hospital-owned practices generated slightly more than half of the median total medical revenue (i.e., cash collections) per FTE physician as physician-owned practices. Median total operating costs per FTE physician followed in a similar pattern, as hospital-owned multispecialty groups reported about half the operating expenses of physician-owned practices, which is reflected in the similar overhead percentages (total operating cost as a percent of total medical revenue). Most importantly, hospital-owned multispecialty practices have half the median total medical revenue after operating costs per FTE physician than their physician-owned peers.

While median total medical revenue after operating costs per FTE physician in the hospital-owned practices is half that of the physician-owned groups, total physician compensation and benefit expenses per FTE physician are similar, due to the “market-based” compensation methodologies used by many hospital systems to recruit and retain doctors.

Having half the net revenue after operating cost per FTE physician and paying market salaries results in hospital-owned multispecialty groups recording significant operating loss per FTE physician — $127,799 in 2015 and $201,703 in 2017 — whereas physician-owned practices essentially broke even both years.

Side-by-side comparison of wRVUs and total wRVUs in the table indicate that physicians in hospital-owned multispecialty practices reported fewer units than those in physician-owned practices. Physicians in the hospital-owned practices reported 89% of the median wRVUs reported by physician-owned multispecialty groups, but only 73% of the total RVUs. The lower value in median wRVUs per FTE physician may be attributed to the greater percentage of primary care doctors in the hospital-owned practices, who perform fewer complex procedures than specialists and therefore report fewer wRVUs per FTE physician.

More importantly, hospital-owned practices report fewer median total RVUs per FTE physician than would be expected based on the difference in wRVUs. This difference is an important factor in explaining the lower revenue in these practices. Producing fewer total RVUs is the result of how hospital systems bill for physician services. “Provider-based billing” allows the health system to create two bills: one for physician services and the other for facility costs. The facility cost payments are revenue to the hospital or system, even though they are not counted as part of the physicians’ contribution to the hospital/system’s bottom line. The impact of provider-based billing can be estimated by examining the Medicare RBRVS fee schedule. In the fee schedule, the total RVU listed for facility billing is reduced anywhere between 30% to 70% of the “nonfacility” amount. Since Medicare payments are strongly correlated to payments by other insurers, we see an overall lessening of total medical revenue in hospital-owned practices than what a physician-owned practice would receive for the same services. 

The Medicare Payment Advisory Commission specifically studied this difference in payment in its 2019 Report to Congress18 when it compared the relative payment for a Level 3 E/M visit for an established patient (CPT® code 99213) in a physician office and in a hospital outpatient department. The total RVUs for the procedure in the private practice is 2.09 compared to the “with facility fee” amount of 1.44 in a hospital outpatient practice. Correspondingly, the payment declines from $75.32 in the private practice to $51.90 when the procedure is performed in a hospital outpatient department. However, the additional facility fee paid through the outpatient prospective payment system (OPPS) to the health system is $115.85. Combining with the outpatient department payment amount, the total payment to the health system is $167.75. The facility fee (which creates the total profit to the system) does not appear on the financial reports of the practice but rather on the books of the system. Therefore, it is possible for the health system to report a loss on the practice even though it earns an overall profit when the facility fee is included.

Revenue cycle information shows that hospital-owned practices report a lower percent of collections from billed charges than physician-owned groups. Table 2 shows that hospital-owned practices are less efficient in collections than physician-owned practices, with about a 1% difference in median adjusted collection percentage, which has the corresponding direct impact of reducing total medical revenue by thousands of dollars per year for equivalent production. More importantly, the payer mix portion of the table shows that while hospital- and physician-owned practices have a similar proportion of payment from Medicare patients, the hospital-owned practices have twice the percentage of Medicaid revenue. The greater percentage of Medicaid payments for hospital-owned practices shows how these practices have a similar payer mix as its parent health system, whereas physician-owned medical groups are better able to determine which patients are accepted by the practice, which allows the practice to maximize the number of patients covered by commercial insurance and lower the percentage of patients with Medicaid. 

Another factor that helps explain why hospital-owned practices report less revenue is revealed in Table 3, which describes the type and the relative proportion of procedures performed by these practices. We see hospital-owned practices report lower median gross charges for medical and surgical procedures performed inside the practice compared to physician-owned practices, reflecting a higher percentage of primary care in hospital-owned practices and, more importantly, the effects of provider-based billing previously described.

Most importantly, the table shows that hospital-owned practices billed hundreds of thousands of dollars less in laboratory and radiology procedures. Rather than a difference in diagnostic and treatment patterns of the physicians in the hospital-owned practices, the lower amount of gross charges reflects these procedures were not billed by the medical group but were performed and billed elsewhere, such as by a centralized imaging center or laboratory within the health system. By centralizing these services, the health system still receives the ancillary services revenue but it is billed by a different entity than the physician group, and the physician group has lower revenue than its private sector peers.

Corroborating the descriptive statistics displayed earlier, Table 4 is a predictive model showing how hospital ownership is negatively correlated with net income per physician at a highly statistically significant level. The model also reveals that wRVUs per FTE physician are statistically significant and positively related to net income per physician, and that operating cost as a percentage of total medical revenue (practice overhead) is, as should be expected, highly correlated with net income. The relationship between payer mix and net income per physician paralleled the descriptive statistics. Since hospital- and physician-owned practices had similar percentages of Medicare, this variable was not at all related to net income, whereas the percentage of Medicaid was negatively related and statistically significant to net income per physician.

The regression shows an Adjusted R2 of .55, implying that the model describes more than half the variance in the total net revenue of these practices. Most importantly, after controlling for the difference in physician productivity, higher operating costs, and the less-remunerative payer mix, the regression model indicates that hospital ownership, by itself, reduces profitability by $126,287 per physician per year, a similar value as the actual median loss of $127,799 per FTE physician reported in the MGMA Cost and Revenue Survey.

Conclusion

Lower physician productivity in hospital-owned multispecialty practices has an impact on their reported profitability. However, it is only a minor factor in the overall explanation of why these hospital-owned practices lose money. Instead, we find that other forces are at work that shift revenues and increase expenses in hospital-owned practices. The descriptive statistics and regression analysis in our research identified the key reasons for these reported losses, specifically:

1. Ancillary services, predominately laboratory and radiology, are shifted away from the practice to the hospital system, reducing the amount of ancillary services revenue reported by the medical group practice.
2. Provider-based billing reduces payment to the hospital-/system-owned practice while simultaneously paying facility fees to the hospital or health system.
3. Hospital-owned practices have twice the percentage of Medicaid payment than private practice with a corresponding reduction in revenue, since Medicaid pays less than commercial payers for similar services.
4. Revenue cycle performance is poorer for hospital-owned groups; the 1% lower adjusted collection percentage on total billings lowers total medical revenue by thousands of dollars and affects the bottom line by a similar amount.
5. With lower total medical revenue per FTE physician and similar total operating expense per FTE physician, hospital-owned practices have higher overhead.
6. These factors, combined with the necessity of paying physicians a competitive, market-based compensation, result in a reported operational loss for hospital-owned practices.

 
It is intriguing that after years of hearing hospital executives blaming employed doctors for diminished physician productivity, we find other factors — payer mix, higher costs and transfer of profitable services — are the primary reasons hospital-owned practices report an operational loss, and that much of the reported operating losses of hospital-owned practices can be attributed to factors unrelated to physician behavior. 

In some sense, the amount of operating loss may not matter to the hospitals or health systems that own physician practices. They acquired these practices not necessarily to make money on the practices themselves, but to increase systemwide revenues while building awareness and loyalty among patients, increase market share, gain leverage with payers and, at least for some health systems, to achieve true clinical integration of services. They may even perceive that operating “losses” are the price to pay for achieving these goals.

Nevertheless, businesses that operate “loss leaders” still want to minimize their overall financial deficits. In this regard, we strongly recommend that hospital and health system executives examine how their practices truly function to gain a clearer picture of the financial state of their physician practices and to realize their owned practices are not as unprofitable as has been alleged. In fact, if all the revenue from the services ordered by the practice are properly attributed to the practice itself, many hospital-owed practices will do more than break even and will contribute a positive margin to their parent organization.

We also strongly recommend that, unless there is evidence to the contrary, hospital executives and health policymakers refrain from bashing their doctors and recognize their and their practices’ contribution to the overall well-being of the larger health system. Instead of complaining, health system executives should congratulate themselves for the foresight of implementing a strategy that integrates physicians into their systems to provide comprehensive, seamless patient care across clinical and inpatient settings. For the health systems that offer truly integrated professional services, their strategy of employing physicians is a wise decision and comes at a price they should willingly pay.

Notes:

1. MGMA. MGMA Cost Survey: 2018 Report Based on 2017 Data. Englewood, CO.
2. Kane, CK. “Updated Data on Physician Practice Arrangements: For the First Time, Fewer Physicians are Owners Than Employees.” AMA. Chicago; AMA Economic and Health Policy Research, May 2019, 2019-3. Available from: bit.ly/34YPzIi.
3. Merritt Hawkins. 2019 Physician Inpatient/Outpatient Revenue Survey. February 2019. Available from: bit.ly/33k1wba.
4. Baker LC, Bundorf MK, Kessler DP. “Vertical integration: Hospital ownership of physician practices is associated with higher prices and spending.” Health Affairs. 2014;33(5):756-63.
5. Cooper Z, Craig S, Gaynor M, Harish N, Krumholz H, Van Reenen J. “Association of financial integration between physicians and hospitals with commercial health care prices. Hospital Prices Grew Substantially Faster Than Physician Prices For Hospital-Based Care In 2007–14.” Health Affairs. 2019;38(2). Available from: bit.ly/3nALNet.
6. The Advisory Board Company. To the Greater Good, Recovering the American Physician Enterprise. Washington, D.C., 1995
7. Ibid.
8. The Advisory Board Company. Stopping the Bleed, Reversing Losses on Owned Practices. Washington, D.C., 1999.
9. Goldsmith, J. The future of medical practice: Creating options for practicing physicians to control their professional destiny. The Physicians Foundation. July 17, 2012. Available from: bit.ly/3qCUkQg.
10. Ermers G, Bundy D. “The challenges of integrating physician group operations.” Dean Dorton. 2014.
11. Kirchhoff, SM. “Physician practices: Background, organization, and market consolidation No. R42880.” Congressional Research Service. Jan. 2, 2013. Available from: bit.ly/3Ibf1IO.
12. Smith, T. “A tale of two owners, Examining losses in physician practices.” MGMA Connection, 2019;(2), 92-94.
13. The concept of the “theory of the business” was created by Peter Drucker. See: Drucker PF. “The theory of the business.” Harvard Business Review. 1994;72(5), 95-104.
14. Hough, DE, Halley MD. “Rethinking the ’theory of the business’ of primary care medical groups.” Group Practice Journal. 1999;48(6), 34-44
15. Conrad DA, Sales A, Liang S, Chaudhuri A, Maynard C, Pieper L, et al. “The impact of financial incentives on physician productivity in medical groups.” Health Services Research. 2002;37(4), 885-906.
16. Gans D. “Urban mythbusting: Correcting the fake news about doctor productivity,” MGMA Connection, 2019;(2), 10-13.
17. MGMA. Annual Cost and Revenue Survey Reports. Englewood, CO.
18. Medicare Payment Advisory Commission. “June 2019 Report to the Congress: Medicare and the Health Care Delivery System.” June 14, 2019; 108-110. Available from: bit.ly/3KmaSU6.