Many hospital and health system C-suite leaders and board members are asking about the size of their organizations’ investments in physician practices, and whether the physician enterprise is optimized to meet organizational goals, now and going forward.

Often, hospital physician networks develop in reactive response to competitive pressures and new opportunities in local markets, as physicians approach organizations seeking employment or partnership. Such development has strengthened and extended hospital networks. However, the growth of subsidies is a major issue for a significant number of organizations.

Every capital outlay and expense requires scrutiny as organizations look to improve quality, outcomes, and patient experience while reducing costs. But a recent survey1 indicates that physician enterprise management and network efficiency—areas that could yield transformative improvements with high levels of investment—are not being addressed at a pace that acknowledges required urgency. Most cost-reduction efforts currently focus on traditional areas, such as revenue cycle, supply chain, and labor productivity.

These circumstances and changing market realities for hospital-employed physicians (see Sidebar 1) suggest the following:

  1. Many hospitals and health systems would benefit from development and implementation of a clear rationale for the number of employed physicians in specific specialties, and a thorough approach to improving efficiency and collaboration within the physician network.
  2. An enterprise-wide approach to cost transformation must include physician operations and related delivery networks.
  3. Physician involvement—if not leadership—in these endeavors is critical to achieving targeted teamwork and efficiencies.

There are four strategies healthcare leaders can put in place to right-type, rightsize, and right-place their employed physician networks and ensure that practices in the networks function effectively and efficiently.

Determine the Right Number and Location of Employed Physicians

The right number of employed physicians can be identified though a disciplined analysis of each service line, the unique position of services offered, the service line’s value proposition, and the future expected value of each business. As services continue to move to ambulatory settings and virtual delivery, hospital volumes are expected to evolve in some areas, so rightsizing the physician network should follow suit. Relevant questions for leadership teams appear in Sidebar 2.

The use of supply/demand modeling to obtain answers is recommended, but reliance on traditional physician-need studies should be avoided. These studies typically are based on the physician-only model of care, in which a primary care doctor or specialist is the sole clinician seeing the patient. They typically point to physician shortages as high as 50 percent in some areas for primary care doctors and more than 65 percent for specialists, and often are used to justify hiring additional physicians.

A better supply/demand model provides insight into the number of physicians actually needed by quantifying the impact of changing patient demographics on the demand side (such as increased consumer expectations related to timely access) and alternative care delivery methods on the supply side (such as use of non-physician professionals).

All delivery methods should be developed with physician input and in a culturally appropriate way that allows clinicians to practice at the top of their license. Thoughtful scenarios can be developed for different staffing models that use advanced practice providers (APPs), such as nurse practitioners and physician assistants, departmental support, and technology to extend the size of patient panels that can be served by the average primary care physician or specialist employed by the organization.


For example, one organization used supply-side modeling to estimate panel size served by an adult primary care medical group under three different pilot care models:

  • Physician only
  • Physician and APP
  • APP (with departmental support)

Figure 1 clearly indicates that the care models using APPs enable larger panel sizes. Panel size with a physician plus an APP nearly doubles (3,000) that of a physician-only model (1,800). As shown in Figure 2, per-patient cost is lower with the models using APPs ($178) than the physician-only model ($248); however, care should be taken to ensure that patients are appropriately assigned to models that align with their specific care needs. Additionally, the piloted care models using APPs demonstrated improved patient satisfaction and access. The full analysis across the organization’s primary care enterprise suggested that costs would decrease by $8.6 million to $9.9 million in salaries and fringe benefit savings alone, as physicians increase their panel count capacity due to improved operational and clinical efficiencies provided by non-physician providers.

Panel size also can be used to inform the mix of primary care physicians and specialists. Over

time, most groups need to evolve to be heavily weighted toward primary care. Leadership teams should think differently about panel sizes and should not assume that historical experience should guide projected levels. Most physician groups are carrying a panel size well below optimized levels.


Supply-side studies that can be used to identify the right number of physicians include:

  • Analysis of the productivity of the existing physician enterprise to determine where increased capacity is required based on overly full schedules or relative value units (RVUs) in excess of reasonable levels
  • Assessment of the organization’s access metrics,  such  as “days to next-available appointment” and “third-available appointment,” along with evaluation of scheduling practices
  • As described more fully in the next section, review of current physician performance to identify high and low performers (physician leadership in such review is critically important)

Finally, financial/affordability analyses should be conducted to identify how much an organization can spend to subsidize the physician enterprise, given system-wide operating results, capital needs, and other factors. Concurrent consideration of capital spending to acquire physician practices, build/retrofit facilities, upgrade information technology, and other major initiatives must be based on solid business plans with uniform criteria that ensure consistent evaluation across opportunities.


Optimize Practice Operations and Staffing

This is a book-length subject, but consideration of a few key issues provides a place to start: benchmark use for clinical improvement, non-physician staffing formulas, and “top-of-license” practice.

Benchmark use for clinical improvement.  An  organization- wide approach to performance improvement is supported by a commitment from the leadership team to aggregate, analyze, and disseminate credible data to clinicians related to quality, outcomes, and cost.2 Benchmark data and advanced analytics using such data enable leaders, quality assurance professionals, and physician/nursing teams to compare clinician performance against historical trends and peer groups. Patterns of performance emerge based on factors such as diagnosis, co-morbidities, treatment type, outcomes, costs, and patient satisfaction. High- performing clinicians can be identified and their practices studied for replication by their peers, as appropriate.

Non-physician staffing formulas. Getting to the right number of non-physician staff is both an art and a science. Traditional non-physician staffing formulas (science) have been based on a per-physician FTE, using median industry benchmarks to identify the number of required staff. A more comprehensive approach (art + science) considers physician productivity. It supports physicians who are working at near or peak capacity with additional staff to meet their unique needs.3 Also important to appropriate staffing is the issue of access to non-physician staff needed by specific populations, such as those with extra acute, chronic, or preventative care needs. Flexibility in staffing formulas is required.

Top-of-license practice. A proven means to avoid unnecessary costs, top-of-license practice involves professionals  working to the full extent of their education and training, avoiding the waste involved in doing procedures or tasks that could be done by someone with less education and training. System benefits from having everyone practice at the top of their license  include improved reimbursement and clear delineation  of tasks, procedures, and duties. Patients also benefit from having clinicians whose skills are continually improving with use.4 To optimize operations, top-of-license practice entails the following:

  • Physicians provide only high-acuity care
  • Moderately acute care is provided by APPs
  • Management of chronic illness is shared by physicians, APPs, and RNs
  • Telemedicine is used as appropriate between providers and directly with patients
  • Pre-visit preparations are performed by support staff to decompress the patient visit time

Dynamic use of facility space under different care models can set the stage for the patient experience while effectively using backstage space to co-locate key processes and create visual cues. Such cues ensure the communication that keeps providers on track with top-of-license duties.


Align Compensation and Performance Goals

Physician compensation that provides fair and stable income while aligning with hospital goals is the single most important factor driving the future performance of a hospital’s physician enterprise. The key principle is to develop uniform compensation standards and metrics that are applied consistently across physicians, locations, and specialties. Standards should cover work effort/productivity, quality, cost-effectiveness, and patient access, and should support the organization’s strategic objectives. Sidebar 3 lists characteristics of better-performing compensation plans.


Of the numerous compensation models used nationwide, compensation per work relative value unit (wRVU) remains the preferred method for a number of reasons:5

  • The physician is held accountable only for his or her productivity and clean charge entry
  • The compensation per wRVU can be adjusted to reflect changes in net collections
  • The method allows for the integration of quality, service, access, and strategic metrics, and rewards  physicians  based on health system goals, such as achieving high levels of patient satisfaction or expense management (e.g., through creation of “shadow wRVUs” paid out in proportion to productivity)

Beginning arrangements for wRVU-based compensation—with incentives for quality, access, service, and other metrics layered on top—typically are 70 to 85 percent productivity based and 15 to 30 percent incentive based.

Physician compensation that provides fair and stable income while aligning with hospital goals is the single most important factor driving the future performance of a hospital’s physician enterprise.

A comparison of physician compensation by specialty with an industry benchmark is recommended. Figure 3 shows percentage variance in compensation by specialty from an industry median based on a conversion factor with compensation per wRVU. Compensation in this organization varies much more than it should, with many physicians receiving less than median compensation (marked by the blue line) and some very expensive outliers receiving 40 to 60 percent more than median compensation. The hospital or health system carries risk associated with those physicians who receive considerably less than median compensation, due to the higher possibility that they may depart for employment elsewhere.


Improve Network Integrity


High network integrity requires a high-functioning referral management process. In such a process, physicians and other clinicians in hospital-owned practices

refer patients to the parent organization’s providers (hospitals, post-acute providers, and other physicians employed by the organization), as appropriate  to patient-care needs. Without such a process supported by high- quality data, analytics, tools, and a thoughtfully designed system to provide appropriate access, patients may seek care directly from an out-of-network provider who is not incentivized or aligned to manage that care.

“Leakage” to non-network providers—whose costs are then attributed to the network—can significantly increase the total cost of care while potentially providing lower quality and less- coordinated care. Goals are to keep care local, where appropriate, and to retain as many of the diagnostic and treatment services as contracts allow. Reasons for leakage include lack of available “in-house” services, medical necessity, patient preference, and physician preference/referral relationships.

An analysis of patient referral flows (or lack thereof) between hospital-owned physician practices and to other owned facilities helps to identify opportunities to reduce internal leakage and enhance alignment with key affiliated provider groups. These findings can inform consumer-facing strategies and the supply- side need analysis mentioned earlier.

Figure 4 illustrates the level of leakage/improvement opportunities by specialty and market for one employed primary care medical group. This group experienced a medium level of leakage with local hematology/oncology referrals (20 to 50 percent of referrals), for example. The group felt that exploration of improvement opportunities might prove productive. Further analysis indicated that patient preference accounted for 40 percent of the leakage, lack of available in-house physicians accounted for 36 percent, while provider preference accounted for 17 percent. Redesigning the patient clinic experience could positively impact patient preferences; extending physician panel size with use of APPs could reduce leakage due to lack of physicians.

Potential Future Scenarios

The current approach to physician employment used by many hospitals and health systems likely is unsustainable in the short or long term. Given this reality, three scenarios can be considered.

First is divestiture and exiting physician practice ownership through sale of the practice to physicians, similar to what occurred in the 1990s. This scenario is unlikely given the tough economics and operational and compliance challenges of private, independent practice.

Second is the sale of practices to for-profit or not-for-profit entities that employ physicians, private equity groups, and others. These entities typically aggregate physicians/groups in regions with the intent to manage as much care as possible locally and direct as much care as appropriate away from hospital settings.

The third scenario, which is recommended, is for hospitals to optimize their physician operations, ensuring the right investment by pursuing the four strategies offered here. Improved data, analytics, and tools to measure the performance of the employed physician operation will enable hospitals and health systems to rightsize and right-locate employed physicians and non-physician professionals, optimize practice operations and staffing, align compensation and performance goals, improve network integrity, and make the necessary adjustments for a more sustainable future.

1Kaufman, Hall & Associates, LLC: 2017 State of Cost Transformation in U.S. Hospitals: An Urgent Call to Accelerate Action. September 2017.
2Morrissey, W.W., Pryor, R.W., and Krishnaswamy, A.: “Using Data and Analytics to Improve Clinical and Financial Performance.” Leadership, Nov. 17, 2016.
3Arnold, C.P.: “A New Approach to Support Staffing in Physician Practices.” Healthcare Cost Containment, April 2017.
4Moawad, H.: “Practicing at the Top of Your License.” MD magazine, May 3, 2017.
5Pizzo, J.J., Sullivan, L., and Ryan, D.L.: “Building the Right Physician Platform” hfm magazine, July 2015.