Seeking a profitable and steady source of patient referrals, in the early 1990s US hospitals and hospital systems began hungrily acquiring primary-care physician practices (Exhibit 1).¹ By 1998, they owned roughly 10 percent of such practices, yet this strategy has done little, if anything, to increase their supply of patients. Moreover, proprietary primary-care practices have become a drain on their parent hospitals, which in 1998 lost a net average of roughly $80,000 per physician from them.² That adds up to more than $1 billion in losses, roughly 30 percent of the net revenue these doctors generated.

Such deals are not easily undone. But by thoughtfully applying two practical ideas from the world of marketing—segmentation and sound channel management—hospitals can increase the number of patients referred to them and turn this channel into a source of competitive advantage.³

A postmortem

Hospital systems overestimated the ability of affiliated physicians to change their referral patterns

Hospitals failed to improve their economic performance through the ownership of primary-care physician practices for several reasons. In the first place, most of the hospitals overestimated the ability of affiliated physicians to change their referral patterns. In a recent study of a large system-owned network of primary-care practices, for example, we found that even several years after the network had been acquired, referrals to it had actually increased by less than 10 percent. In absolute terms, fewer than half of the patients seen by the system’s primary-care physicians received follow-up care from affiliated specialists or hospitals.

About 75 of 100 physicians in recently acquired primary-care practices told us that there were four main barriers to changing their referral patterns. Often, neither the doctors nor their patients were familiar with the affiliated specialists. Patients were unwilling to travel unusually long distances to see them. Payor incentives or policies sometimes encouraged primary-care physicians to refer patients to unaffiliated hospitals and physicians. And affiliated hospitals suffered from shortages of critical technologies (such as CAT scanners) or surgical facilities, thus causing excessively long waits for patients and physicians alike.

The second reason hospital systems could not improve their economic performance by purchasing practices of primary-care physicians was their new mode of compensation. Traditionally, primary-care physicians ran their own businesses and were paid for each patient they saw. Because every dollar they saved went straight into their own bank accounts, they had every incentive to keep their support staffs lean, to control other expenses, and to monitor billings and reimbursements closely. But hospitals encouraged primary-care physicians to join hospital systems by offering lump-sum payments over and above the value of the assets of the practices. The hospitals then put the physicians on salary. Assured of a certain income regardless of performance, the physicians’ productivity declined by 15 to 20 percent in the two years after acquisition.

Finally, hospitals tended to overestimate the value of direct referrals from primary-care physicians into hospital systems—referrals that actually generate relatively little income (Exhibit 2). More than three-quarters of an average hospital’s net income is derived from referrals by specialists. Specialists refer not only more patients than primary-care physicians do but also patients requiring more elaborate care—thus generating higher hospital bills and, potentially, more profit. Further enhancing the specialists’ importance is the recent tendency of more informed and assertive patients to bypass primary-care physicians and go directly to specialists.

How to stanch the bleeding

Far from raising the profits of hospitals, the ownership of primary-care physician practices has clearly created an additional economic burden for them. Equally clear is the essential role specialists play in influencing the flow of patients and driving the economics of hospital systems. Moreover, it is telling that three of the four reasons primary-care physicians gave for declining to make referrals within their own hospital systems concerned their doubts about the quality, convenience, and availability of the services those systems offered patients and physicians. A successful channel management strategy must therefore acknowledge both the role of specialists and the importance of excellent service.

Segmentation

Good channel management begins with the segmentation of channel intermediaries, in this case primary-care physicians (independent or employed by the hospital) and specialists. Primary-care and specialist practices should be segmented according to their current and potential economic contribution to the system.

For practices owned by hospitals, the current contribution is the operating profit of the practice, plus any economic contribution its patient referrals (either to independent but affiliated specialists or directly to the hospital) make to the system. The current contribution of nonemployed practices excludes any consideration of their operating economics and is simply their contribution from referrals. The potential economic contribution of a practice is its ability to increase the number of referrals it makes to the system—an ability influenced by the local reputation of the practice, the number and quality of competing physicians and hospitals in the region, and the hospital’s location, capabilities, and reputation.

Segmentation can help a hospital system devise appropriate strategies to improve the economic contribution of most practices. As Exhibit 3 shows, practices can be divided among four segments. Practices in the upper left quadrant ("grow the patient base") make a significant economic contribution to the hospital but have only a limited ability to refer more patients to it because of its distance from patients (compared with other hospitals), the number of competing practices, or constraints imposed by regional health plans. Members of a general-surgery practice in a rural area, for example, might have a hard time persuading patients of the need to bypass a local hospital for one farther away. Little can be gained from investing in efforts to increase the percentage of patients these practices refer to the parent system. What hospitals can do is help them enlarge their patient bases to increase the absolute number of referrals they make.

Practices in the upper right quadrant ("lock in referrals") also make a significant economic contribution to the affiliated hospital system but find it easier than practices in the upper left quadrant to shift their patient referral patterns. A cardiothoracic-surgery practice in a large urban area, for example, would probably fall into this category. Hospitals ought to manage such practices with a view to "locking in" their current referrals by creating barriers that make it harder for them to shift their patients to alternative hospitals. A secondary objective should be to help the practices enlarge their referral bases.

Practices in the lower right quadrant ("increase percentage of referrals") make only a small current economic contribution to the system but have the ability to shift their patient referral patterns. Before locking in their referrals or helping such practices enlarge their patient bases, a hospital should take measures to increase the percentage of patients they refer to the hospital.

As for practices in the lower left quadrant ("take no action or restructure relationship"), they not only make a minimal contribution to the system but also have a limited ability to redirect the flow of patients—for the same reasons that make it hard for practices in the top left quadrant to do so. Examples might include small, rural primary-care practices. If these practices are independent, a hospital need take no action, since they are not an economic burden. If the hospital owns them, it should change the relationship to improve the economics, perhaps by selling them or altering the way they are compensated.

Many primary-care practices fall in the lower left quadrant of the matrix. Most specialist practices fall in the quadrants on the right side.

Four levers

McKinsey’s channel management work across a multitude of industries has identified four key levers for enhancing performance: developing business skills, supporting infrastructure, improving coordination among channel players, and devising incentives. Applying these levers to the appropriate segments can help create an environment where practices are more successful and increase their contribution to the system (Exhibit 4).

Skills. Practices can grow by developing better basic marketing and business skills, which even the most productive physicians and practices often have little time to improve. These skills are most valuable to practices in the top quadrants of the matrix.

Hospitals can help practices grow and become more productive by raising their skills in operations, marketing, services, and strategy

Hospital systems can help practices grow and become more productive by improving their skills in four areas: strategy, marketing, services, and operations. Strategy might, for instance, encompass the ability to understand the potential of the local market, the competitive forces at work there, and the way patients choose physicians. Marketing efforts could include co-branding practices with their affiliated hospitals or devising advertisements to recruit physicians. Service development might involve efforts to identify new offerings, such as telemedicine or extended hours for the convenience of working patients. Operational skills include claims management, purchasing, and the automation of the physicians’ key back-office processes.

Infrastructure. Practices of physicians are usually small businesses with little access to significant capital for investments. If hospitals could help practices buy imaging or diagnostic equipment, acquire office space, or establish nearby testing facilities, those practices could attract more patients. Spending on infrastructure would therefore not only help practices in the upper two quadrants increase their market share but also lock in referrals from practices in the top-right quadrant by creating switching costs for physicians.

Coordination. Better coordination is a useful lever for all physician practices but particularly for those in the two right quadrants of the matrix. Since referrals to specialists can slow to a trickle if the specialists don’t continually update the referring physicians on the status of patients, hospitals should develop information systems to track and report their patients’ progress throughout the entire care process. They can also improve coordination between specialists and primary-care physicians by, for example, investing in systems to inform primary-care physicians about the capabilities of affiliated specialists and in call centers to help primary-care physicians choose the right one.

Incentives. Particularly in the case of practices in the two right quadrants of the matrix, incentives are an important lever for maintaining and increasing the number of patients they refer to a hospital. Although federal regulations rightly prohibit hospital systems from paying practices for referrals, other kinds of incentives are allowed. Three in particular deserve mention.

First, our work with hundreds of physicians has impressed upon us the importance they place on the accessibility, efficiency, and quality of hospital services such as admitting, hotel services, and operating rooms, as well as clinical support services, including laboratories, imaging, and pharmacies. This suggests that enhanced customer service—both for patients and physicians—will not only maintain current referral levels but also raise them.

Second, whenever possible, hospitals should link the compensation of physicians to their productivity. For employed physicians, this would mean measuring the outcomes and efficiency of the care their patients receive. For independent physicians, it might mean forming partnerships—perhaps through jointly owned clinics or surgery centers—that would give them an incentive to maximize their own economic performance.

Third, to make it harder for practices to shift their referrals elsewhere, hospitals should, for example, consolidate purchasing for practices, handle their back-office work, and integrate their information systems.

The damage already wrought by the ill-advised acquisition and subsequent mismanagement of primary-care practices can’t be undone, but it can be mitigated in the present and avoided in the future. The segmentation of physician practices—primary care and specialist alike—along the lines described here can help hospitals turn this channel into a source of competitive advantage.