Cohen, Jon R. MD; Dowling, Michael; Gallagher, John S. T.
In 1997, after a nine-month legal challenge by the United States Department of Justice Antitrust Division, the North Shore Health System and the Long Island Jewish Medical Center merged to become the third largest not-for-profit, academic health care system in the United States, with revenues in excess of $3.2 billion. This health care system now includes 14 hospitals, two skilled nursing facilities, three regional trauma centers, 88 ambulatory care centers, 4,700 hospital beds, 6,600 physicians, 7,440 registered nurses, 3,300 volunteers, 1,000 research studies, 1,140 community service programs, 750 full-time faculty, and 1,100 residents in 55 approved training programs. It provides health care for 265,000 children and 700,000 elderly per year. By the end of the year 2001, over 200,000 patients will have been admitted to the system's hospitals, 130,000 people will have undergone operative procedures, 2,000,000 people will have been seen in outpatient clinics, nurses will have visited 574,000 homes, and 22,000 babies will have been born in the system's hospitals. Three years after its inception, the North Shore—Long Island Jewish Health System has emerged as one of the more successful academic hospital system mergers in the country. How was this achieved?
In this review we analyze the specific factors responsible for the relative success of our clinical merger between the two large tertiary academic hospital systems, review our failures and the lessons we have learned, and discuss the future challenges of our clinical integration strategy. We do not address the nonclinical operational aspects of our corporate merger activity in system departments such as legal affairs, purchasing, external affairs/development, managed care contracting, finance, quality assurance, human resources, biomedical, materials management, pharmacy, medical supplies, plant operations, security, and laboratories. Merging these departments was easier and occurred more quickly than did the integration of the clinical departments. Single-leadership appointments and the merger of these nonclinical departments provided economies of scale without loss of market share.
Commitments at the Onset
The strategic decision and commitment for the merger between the two hospital systems arose from and were directed by the two boards of trustees. Several of the reasons that contributed to the decision to merge included significant decreases in reimbursement from managed care companies, lack of funding for depreciation from managed care companies, significant decreases in graduate medical education funding as a result of the Balanced Budget Act, and a strong sense that the whole resulting from the clinical merger of these two systems would be greater than the sum of the individual entities. One and a half miles apart, the two 800-plus-bed tertiary medical centers had been arch rivals and competitors for over 40 years. The intent of the merger of these two systems was to create savings that could then be reinvested in providing better health care for the communities that they served. Savings would occur by consolidating corporate services, infrastructure, support services, and the clinical enterprise; preserving assets; fostering economies in purchasing, professional services, information systems, and physician support services; and avoiding some capital out-lays.
The executive leadership was responsible for designing and subsequently implementing the vision and future of the newly merged entity. It was understood from the beginning that for the merger to succeed, the administrative and medical leaders had to be willing to set aside parochial interests and convince the rest of the organizations that the integrated health system would be greater than the sum of its individual parts. As in most health care mergers, convincing the clinical chairpersons and senior hospital site administrators at the two tertiary institutions of the benefits of the newly merged system was the most difficult task. The small group of leaders at both institutions that planned the merger anticipated that the chairpersons and division chiefs would instinctively protect their individual empires in efforts to avoid change and maintain control. Site administrators would be appropriately hesitant to change any program or department in a way that might have a negative impact on their revenue streams. A small but significant group of administrative and clinical leaders had to “buy in” that the clinical merger was the right vision for the future of the organizations. Those of us involved in fostering the merger learned that to achieve buy-in and make clinical integration a success, communication, compromise, and time in getting to know one another were, and continue to be, the most important elements. We soon became aware that the process could not be rushed, since, in fact, the most important contribution of these elements was the development of trust among the senior administrative and clinical leaders.
The most effective mechanism to achieve this trust was the provision of multiple common platforms for discussions by the leadership group, while simultaneously keeping local meetings at the tertiary sites to a minimum. Our most important meetings, which are ongoing, are (1) meetings of the Joint Conference for Professional Affairs, where the medical leaders from the entire merged health care system and the board of trustees meet to discuss quality initiatives across the system; (2) a biweekly meeting of all the chair-persons and the senior administrative staff members from both tertiary institutions, which provides a forum for open discussion of issues affecting both tertiary hospitals; (3) a weekly meeting of the chief executive officer, the chief operating officer, the chief medical officer, and tertiary site executive administrators to discuss ongoing clinical issues relative to the two tertiary campuses; and (4) a monthly meeting of the chief executive officer, the chief operating officer, and the chief medical officer with the clinical leaders from all the hospitals in the merged system.
Rules of Engagement
At the very outset of the merger, there were extensive discussions about clinical integration, how to rationalize resources, and the plan for program and clinical consolidation as a way of reducing costs. Our consultants estimated at the time of the merger that $49 million would be realized in cost savings through the opportunities created by clinical consolidation. The assumption was that merging the clinical programs at the two tertiary hospitals would result in significant economies of scale while simultaneously allowing those hospitals to retain the full range of specialization and support functions. As a merged entity the two tertiary hospitals could theoretically consolidate programmatic leadership and support functions as well as eliminate duplicative staffing. The guiding principles were (1) to have a single chairperson within each department, with associate chairs for the larger departments; (2) to have one chief of each division or section within each department; (3) to reduce the number of residency coordinators; (4) to reduce by 20% the number of full-time faculty; and (5) to carry out associated reductions in the numbers of administrative and clerical staff commensurate with the above reductions.
Nothing could have been further from reality. Almost none of these guiding principles for clinical consolidation were implemented or ever realized. Almost none of the projected savings were realized; neither was the bulk of the strategy to decrease staffing accomplished. Having observed several failed attempts at clinical integration across the country, whereby large numbers of faculty and programs abandoned their institutions, and drawing upon our own limited experience with unsuccessful attempts at forced clinical integration, we evolved a new set of guiding principles. These new “rules of engagement” became a very pragmatic set of principles to move the merger forward without breaking it apart.
The guiding principle for clinical integration became one of not merging or integrating any clinical service for the sole reason of merging, but rather to encourage clinical integration where and when it made sense to achieve specific goals such as to improve residency training; to enhance research opportunities; to refine the quality of care; to develop new programs or expand existing programs; to develop economies of scale; or to take the opportunity to improve clinical leadership. “Consolidation” was no longer discussed, as it was usually interpreted as downsizing of a clinical service at one tertiary site or the other.
Consideration for clinical integration of any department, division, or program must include an analysis of the relationship between the clinical departments and their hospitals. These discussions must include input from senior hospital administrators to determine the advantages or disadvantages of departmental and divisional integration. No program would be moved from one tertiary site to another without considering the impact on the other. In fact, three years into the merger, no program has been moved or discontinued at either tertiary facility. Early on, we recognized that each medical center had formed its own “ecosystem” whereby most departments were dependent on the others for the optimal functioning of a full-service tertiary center. The concept evolved of “building the twin towers” to maintain the highest possible degree of quality and service at both tertiary hospitals for the delivery of health care to the communities that they served and to maintain and expand the market shares of both tertiary centers. Although only they are 1.5 miles apart, the tertiary centers have specifically different market shares and name recognition in many clinical areas. The impact of removing any critical service at a full-service hospital was deemed dangerous to the long-term survival of that institution and not in the best interest of the community it served.
As an example of the twin-towers approach, early in the merger, an enticing opportunity arose to merge the two hospitals' divisions of cardiac surgery and consolidate the program at one institution. It seemed like a great opportunity. One site performed 1,200 open-heart sugeries annually; the other performed 600 annually. A combined program would have created the largest cardiac surgery program in New York State. Furthermore, the economies of scale seemed enormous. However, after careful analysis it became obvious that removal of cardiac surgery from one of the facilities would transform the entire character and mission of that hospital. Eliminating cardiac surgery would have had a significant impact on the training programs in cardiac surgery, general surgery, cardiology, medicine, and rehabilitation. It could possibly have had an enormous impact on emergency services, consultation services, intensive care units, cardiac care units, and the level of nurses' training. Furthermore, it would have affected the referral base from the voluntary staff, since the perception of any decrease in services would weaken the reputation of the institution and the subsequent ability to draw patients across the spectrum of diseases. In the final analysis, the impact of removing the cardiac surgery division of one of the hospitals would have been a marked decrease across the board in the volume of discharges at that institution.
Two specific exceptions to the twin-towers policy arose. These concerned our children's hospital and psychiatric hospital. An important part of our strategy is to continue to support these two hospitals while maintaining the appropriate levels of pediatric and psychiatric services at the two tertiary hospitals. In addition, we are currently considering merging pediatric cardiac surgery and pediatric bone marrow transplantation to one site.
Clinical Integration, the Medical Schools, and Academic Appointments
The North Shore University Hospital faculty are appointed at New York University School of Medicine, and the Long Island Jewish Medical Center faculty are appointed at the Albert Einstein College of Medicine. At least three other medical schools have long-standing relationships with hospitals in the health care system, in each case pertaining to residency-related issues. It became apparent that any clinical integration would be successful only if we were to avoid addressing the issue of academic appointments. Both institutions have strong ties to their medical schools and have third- and fourth-year medical students rotating on almost all of the clinical services at all times. It would have vastly complicated the clinical integration of the two campuses to have the medical schools participate in clinical integration decisions. However, we knew it was extremely important to maintain close ties with both medical schools in anticipation of appointments for future faculty, teaching medical students, graduate medical education, and cooperative research efforts. One of the sensitive issues was the decision as to which institution would be the sponsoring one for the different residency or fellowship programs when programs were merged. Albert Einstein College of Medicine was and remains the sponsoring institution for the Long Island Jewish residencies and fellowship programs, just as the New York University Medical School was and remains the sponsoring institution for such programs at the North Shore University Hospital. To circumvent these issues, the decision was made to maintain and continue to develop the graduate medical education consortium whereby the North Shore—Long Island Jewish Health System could become the parent sponsor for residency programs as needed.
Faculty Practice Plan
As we began to consider any significant integration of clinical departments or divisions, the faculty practice plans at both institutions needed revision in order to be represented on a common platform. The two plans possessed similar revenue bases, with similar numbers of faculty; $150 million in revenues at North Shore University Hospital with 325 full-time faculty, and $160 million in revenues at Long Island Jewish Medical Center with 425 full-time faculty. Although the plans were similar in size, the details of their governing structures, histories of hospital support to the individual clinical departments, and billing infrastructures were significantly different.
A senior administrative and clinical leadership group composed of leaders at both hospitals evaluated both plans, and over a six-month period of time developed a new over-arching faculty practice plan. This effort had two significant results. First, it communicated to the rest of the organization at each hospital that the commitment to the merger was so strong that the leaders were willing to change the faculty practice plans at both tertiary institutions. Second, it provided another very important vehicle for extensive communication and serious discussions between the leaders of the two tertiary institutions. The focus of the committee's work was to reconfirm its mission to establish a faculty practice for the merged institutions, develop consistent guiding principles for hospital support, and develop a “best practice” billing infrastructure.
The committee reaffirmed its mission and the new practice plan's organizational structure by reviewing several faculty practice models that existed in similar academic medical centers across the country. This review resulted in the development of a single mission and vision statement for the new plan, the development of a governing board with equal representation from both tertiary sites to oversee the plan, and a structure committed to strengthening the close relationship of the plan to the hospital.
The committee's review of the levels of hospital support provided to the different departments showed appreciable differences in the “deals” between the two hospitals and the manners in which individual departments were treated within the same hospital. Major areas of focus included the levels of reimbursement for the teaching and administration of residents, reimbursements for physicians' time spent on administrative or clinical services, and the charges for typical practice expenses such as overhead, malpractice, billing, and the mission-related tax (i.e., dean's tax). Working out approaches to those areas meant, among other things, that the two organizations had to “open their books” to one another for review. There is almost nothing more serious or sensitive than when organizations take that revealing step. Pursuing the major areas of focus mentioned above provided us with an opportunity to objectively compare departments at the two tertiary sites. Accusations across the campuses of financially “hemorrhaging” departments and “poor productivity” by certain faculty on one campus or the other could now be objectively measured. The committee's work, as just described, was a major step forward in evaluating the departments equitably.
With a combined net revenue base from patient care of over $300 million, the tracking of and timely and accurate collection of dollars are extremely important to the chairpersons and faculty. This issue alone became a common ground for agreement across both campuses. An extensive review of both faculty practice plans' structures and an evaluation of outside billing vendors resulted in the design of an in-house operation that has both a centralized component and a decentralized component to its operation.
Funding of Research
Another very important observation that was gained from the analysis associated with creating the new faculty practice plan was that the funding of research was vastly different at the two institutions and the difference would have to be addressed to make the new faculty practice plan functional. Issues related to departmental funding of research, outside funding, overhead arrangements, indirect costs, and the development of other sources of research funding spearheaded many of the discussions of research funding. These issues were resolved and the future of research within the health system was assured by the development of a separate research institute with its own board and research director reporting back to the parent health care system.
Models for departmental leadership
Four different models for departmental leadership evolved through discussions and trial and error:
One chairperson for both campuses: In this model the chairperson is responsible either for the one department across both campuses or for the department on each campus, and frequently designates senior faculty members for different areas of responsibility, such as residents' education, quality assurance, faculty practice issues, and continuing medical education.
One chairperson for both campuses and a senior-level vice chairperson at each campus: In this model the vice chairperson serves as the senior physician at each campus for daily operational issues, reporting up to the chairperson.
One chairperson, with a senior level physician as a vice chairperson: In this model the chairperson has his or her base of clinical operation at one campus, and the vice chairperson has his or her base of clinical operation at the other campus. The vice chairperson serves as the clinical leader at one campus, with responsibilities for the day-to-day activities at that campus, and reports directly to the chairperson for major issues. The chairperson has responsibility for day-to-day operational issues at the other campus. An example of this is the chair of surgery; his clinical practice is at one campus but he oversees the entire department across both campuses. The vice chairperson has operational responsibilities for the operating rooms at the other campus and functions as the clinical leader for day-to-day issues, reporting to the chairperson.
Two chairpersons, with one at each campus.
How departmental leadership was established
Opportunities for changes in leadership presented themselves by natural attrition arising from retirement, death, or a chairperson's move to a position outside the system. Opportunities for new leadership were created in several instances by developing new departments; for example, the development of the Department of Laboratory Medicine as a separate department from the parent Department of Pathology at both institutions, and the creation of a Department of Neurosurgery, which had been a division in the Department of Surgery at both tertiary institutions. In some arrangements, such as that for orthopedics, the speciality was represented by a department at one institution and a division at the other. In this instance, orthopedics was given departmental status at both sites and a new chairperson was appointed. In some circumstances, a proactive decision was made to change the leadership.
Selecting and appointing new chairpersons
Certainly one of the most passionate and potentially contentious decisions surrounding a merger is the selection of new leaders. Two broad processes evolved: either directed appointments with no search, or a formal search process. A directed appointment always consisted of having an existing chairperson at one campus become the leader at both institutions or having an existing division chief become the chairperson of the new department. No directed appointment was made from outside the organization. The decision for a directed appointment requires a significant amount of “buy-in,” with approval of the board of trustees, senior administrative leadership, and senior clinical leadership. In our circumstance, it required formal approval from the medical leadership and the board of trustees to waive the search process. Appointing a leader from within the organization from one of the tertiary institutions to be the leader at both campuses was almost always fraught with a significant amount of paranoia from the “receiving” faculty and the voluntary staff. It is during these times that the medical staff feel as though they are “being taken over” by the other institution; they resent senior administrative and clinical leadership for having “sold them out” in the interest of a smooth merger.
Resolution of these issues requires a significant degree of constant communication and leadership from the chosen individual. In the instances where a formal search was performed, it was critical to have had equal and adequate senior clinical representation on the search committees from both the full-time staff and the private staff at both institutions. Although in the past the trustees had not directly participated as members of the search committees, the sensitive issues involved made it important to appoint board representatives to the committees to help further gain the confidence of both institutions that the selection of the new chairperson would be in the best interest of all involved.
Current status of clinical leadership
Given the four different leadership models, the different opportunities afforded us to choose leaders, and the different ways we selected leaders, we arrived at our current leadership status by nine different scenarios:
▪ One chairperson with two vice chairpersons, with the chairperson being a new leader recruited from the outside;
▪ one chairperson with one vice chairperson, with the new leader recruited from the outside;
▪ two chairpersons, both existing leaders;
▪ two chairpersons, one existing chairperson and the other, formerly a division chief, appointed as a chairperson;
▪ one chairperson, an existing leader where the other leader had retired;
▪ one chairperson, an existing leader where the new department was created at the other tertiary campus;
▪ one chairperson of a new department created at both campuses, with the leader recruited from the outside;
▪ one chairperson of a new department created from existing departments, with the leader recruited from the outside; and
▪ one chairperson on one campus and a division chief for the specialty on the other campus, with no decision to date as to which model is more appropriate.
Current Levels of Clinical Integration
One of our chairpersons, who is somewhat skeptical about clinical integration, likened it to fine wine; “You can have a great red wine and a great white wine; the minute you mix these two great wines you have a poor result.” As the nine different scenarios for clinical leadership evolved, the chairpersons were encouraged to clinically integrate as much as possible and as they felt was appropriate when the opportunities presented themselves. Clinical integration was not forced but was allowed to evolve naturally over time with help and encouragement from the concerned parties. The intention was not to have the chairs create larger but mediocre departments or divisions by forcing integration at the expense of two good departments or divisions functioning well independently.
The broad areas for integration were conferences, residency training, sharing of faculty and programs, financial sharing, and research. The current level of clinical integration was determined by surveying the chairpersons in each of these broad areas of clinical integration. For each department, each of those areas was then assigned a point value to determine the comparative level of integration (zero points, one point for minimal integration, two points for moderate, and three points for full integration). (See Table 1 for full definitions of these three areas.)
All 19 departments were surveyed in the year 2000 (see Figure 1). Three departments had no integration at that time, but represented different organizational models at the two institutions. One of these departments was a private practice group at one institution and a full-time group at the other tertiary site, and neither had a residency program. Two departments were full-time departments with approved residency programs at one tertiary site and divisions of a major department with a private practice model at the other tertiary site. No recommendation had been established to address the leadership model in any of these three departments; thus, little or no clinical integration had occurred.
The extents of clinical integration of the other 16 departments ranged from 20% to 73% (see Figure 1). Five departments each had more than 50% integration. The overall integration for these 16 departments taken as a whole was 42%. The overall integration for all 19 departments, considering the three departments with no integration, was 35%.
In comparing the total integration for the five different broad areas of departmental activities, the greatest degree of integration has occurred with respect to conferences (60%) and the least in regard to finances (27%—see Figure 2).
We also analyzed the leadership models in relation to the degrees of departmental clinical integration (see Figure 3). The highest average level of integration (67%) occurred with dual leadership (co-chairs) in two departments where the chairs had had long-standing, amicable, professional relationships prior to the merger. Nine departments with single chairpersons had an average of 46% integration. The lowest average level of integration (16%) occurred for two departments, each of which had two chairpersons who had little or no relationship prior to the merger.
The levels of integration described above are, in general, increasing several months later as this article goes to press.
Three years after the inception of the merger, clinical integration for the 16 departments where we encouraged integration was an average of 42%. A recent study of 40 integrated delivery health care systems that were considered successful by their peer organizations showed that not one of these organizations regarded itself as a fully clinically integrated system.1 We anticipate that it will take at least two more years for us to achieve 80% overall clinical integration. Part of the reason for this time frame is that several departments have just recently established unified, single leadership. However, we think that there is no reason to establish an absolute level of “acceptable” integration, inasmuch as we are uncertain what overall level of integration would be appropriate. In fact, 100% integration is probably not achievable, nor should it be the end goal. In reviewing the five areas for clinical integration, it is not surprising that integration has occurred most frequently with respect to cross-campus conferences, since it is the easiest and least contentious departmental activity for the two staffs. As we suspected, the financial area has had the least amount of integration, and will probably be the last to be significantly integrated across the campuses.
The two most integrated departments function under the co-chairperson model, where these co-leaders had had amicable personal and professional relationships for over 15 years predating the merger. Nine departments had single chairpersons, with an average level of 46% integration that had been achieved over two and a half years. We believe integration will continue to increase significantly over the next three years under the single-chairperson model. It was predictable that clinical integration was least where there were two leaders who had had no established relationship before the merger. An effective co-leadership model in these circumstances has had little success. There has been no integration whatsoever in several departments because we have not yet been able to decide upon the leadership models for these departments for the future. Some are of the opinion that the administrative leadership should mandate a single leader in these co-leader and unresolved departments and move forward with clinical integration.2 We continue to believe that this approach is inappropriate and leads only to further physician disenfranchisement and possibly the loss of talented physicians from those departments.
Our failure to recognize and acknowledge specific clinical issues parochial to one campus or the other frequently caused a great deal of backlash from both full-time staff and private staff. Our attempt to impose any clinical integration strategy without adequate “vetting” on either campus was almost always met with significant resistance. As with most issues, a comprehensive discussion among all parties was mandatory to move any agenda item forward. One cannot overestimate the “human” factor, especially pride in a program, department, or division, when discussing clinical integration issues. We also failed early in the merger effort to prevent some loyal physicians from leaving the tertiary hospitals in fear of the unknown impact of the merger on their clinical practices. Better communication and discussion with these physicians probably would have prevented the loss of some of these valuable practitioners.
In reviewing our experience to date, the common denominators responsible for the relative success of our clinical merger have been early recognition of the above-mentioned failures and subsequent constant communication among the leadership and staff, flexibility in building leadership models, and patience in letting events progress over time in order to develop an atmosphere of trust within the organization. Another critically important factor for successful clinical integration is the existence of a senior executive structure that the organization clearly recognizes as empowered to make decisions. In our case, the chief executive officer, chief operating officer, and chief medical officer constitute the clear central authoritative body responsible for all financial and clinical decisions.
The participation of private and community-based physicians in the entire process, with consideration of their cultural differences and practice patterns relative to each tertiary hospital, is also of paramount importance to the success of any clinical integration. Interestingly enough, 75% of health care mergers do not succeed because of problems with cultural issues that surface when clinical integration is attempted.3 In fact, in 1999 several large academic health care system mergers failed, including the merger of Stanford University Hospital with the University of California at San Francisco and the merger of the Geisinger Health System with the Milton S. Hershey Medical Center. The failure of so many health care system mergers has resulted in a significant decrease in new hospital merger and acquisition activity. In 1999 there was a 30% decrease, with 111 hospital mergers and acquisitions, compared with 144 such deals in 1998 and 197 in 1997. Among these mergers and acquisitions, not-for-profit hospitals continue to dominate the market.4
The measure of success for a merged, integrated delivery system will continue to be its ability to operate as a seamless network across the system. In a successfully clinically integrated system, care guidelines and protocols will help to enforce one consistent standard of care. If done correctly, clinical integration will help to coordinate this consistent standard of care across the system and help to define better patient outcomes.1,5