Like other U.S. medical schools in the 1990s, the University of Minnesota Medical School encountered new challenges that included increased competition for health care services and for research funding, more limited state support, and greater demands for financial accountability.1 Yet at the medical school, these demands for accountability occurred within a very decentralized structure, where each department's financial operations were autonomous. During that period, the medical school also experienced its own particular challenges, including incidents of financial mismanagement and intense fragmentation of faculty practice plans.
The Roots of Change
In a 1993 Deloitte & Touche Management Consulting report to the University of Minnesota, it was noted that internal audits of the medical school had resulted in citations of improper financial activities, “including purchase of personal property with university funds, falsification of overtime reporting, and sale of university property for personal gain.” The risks of these improper activities would continue, the analysts wrote, unless the medical school strengthened its administrative systems. One of their suggested solutions was to cluster departments under administrative centers that would consistently apply policies, standardize reporting for consistency, improve service, and centralize oversight through a position in the dean's office.
At that time, a group of medical school faculty leaders, including most clinical department heads, submitted a formal letter to the university president arguing against administrative centers. They wrote that department heads must have control over individual department finances to pursue their missions. So, the medical school tightened controls a different way. A new dean's office position for an associate dean for administration was established, and the person holding that position was charged with hiring new staff who would be responsible for finance, human resources, physical resources, information technology, and practice plans schoolwide.
Even though the administrative centers temporarily were off the agenda, another of Deloitte & Touche's recommendations about the practice plans received attention and provoked change. The auditors reported that one of the continuing sources of financial risk for the medical school was the private practice plans, which numbered as many as 40 for the 18 clinical departments at the time. The practice plans were vulnerable to financial problems because of a lack of consistency in accounting practices across departments and a lack of oversight by the medical school. In response, the University of Minnesota regents expressed concern about these private practice corporations existing outside the purview of the university. The regents said that they would prefer, but would not mandate, that the medical school have only one practice plan. They asked that a practice plan administrator position be established in the dean's office to improve the university leadership's knowledge of the inner workings of those practice plans.
In addition, the regents asked that each medical school department have only one practice group. To address the request, the faculty and clinical leadership expended a great deal of effort and, by July 1995, had accomplished the first consolidation in which each department had only one practice plan. The medical school faculty then independently came to the conclusion that there would be competitive advantages to being unified into one practice group encompassing faculty from all departments. They pursued that goal. The single faculty practice plan was founded as an independent nonprofit organization in 1997 and was called University of Minnesota Physicians.
While the practice plans were being reorganized, the University of Minnesota decided to sell its hospital to cope with financial exigencies. Negotiations between the university and Fairview Health Services lasted a year and a half and resulted in a new kind of partnership between an academic medical institution and an outside not-for-profit health care entity. The sale of the University Hospital to Fairview Health Services was completed in 1997. The hospital is now known as University of Minnesota Medical Center, Fairview, and is the primary practice site for members of University of Minnesota Physicians.
Developing a strategic plan
After the sale of the hospital in 1997, the medical school dean at the time, Alfred Michael, commenced a strategic planning process to address the following question: Now that the school is structured differently, with one unified faculty practice group, should it be managed differently? At a planning retreat, a facilitated discussion addressing how to move the school forward revealed that administrative costs, policies, and streamlining to increase efficiency were secondary, yet very important, to the school's goals. From the point of view of faculty, the medical school was spending too much on administration and not enough on advancing the school's educational and research goals. Furthermore, faculty believed that the administrative services provided were not of a consistent high quality. The dean had not intended to change administration, yet this discussion showed that this issue needed attention and reform. As a result, the dean set up a medical school administrative structure work group to analyze the school's administrative structure and to deliver recommendations. The initial goal of this analysis was simply to improve administrative services without increasing expenditures.
As resources for initiating administrative changes, the work group not only had the 1993 Deloitte & Touche report's recommendations of administrative centers, but they also had an example of a new administrative approach emerging within the medical school. The department of medicine administrator, when she heard that the smaller and financially troubled department of dermatology would be gaining a new head, recognized an opportunity for that smaller department to share limited resources, to save money by reducing the number of administrators, and to improve service for dermatology faculty. She proposed to the department heads and to the dean of the medical school that the two departments share administrative services and that the staff be reorganized to “deliver added value at a lower cost” (M. Christensen, personal communication, 2005). She also promised to those administrators affected by the proposed reorganization that no employees would lose their jobs. (She was able to keep this promise, though some shifting of responsibilities did occur.)
This new interdepartmental model of sharing administrative responsibilities started as a test with the departments of medicine and dermatology, and, within a year, a third moderate-sized department, neurology, had joined medicine and dermatology in the medical school's new clustered approach to departmental administration. The administrator who initiated this change, interviewed several years later, said there were more barriers in the process than she had anticipated. One barrier was presented by underlying issues unrelated to the formation of the administrative center that initially detracted from the faculty's enthusiasm for the center. To resolve these issues, the administrative center director and staff immediately sought to prove their value to the faculty. They addressed long-standing issues among faculty, increasing their salaries and improving their retirement packages.
Still, in light of the financial difficulties the departments experienced in the mid-1990s, departments needed to change, and department heads looked for ways to cut costs. This shared administrative structure pioneered by the departments of medicine, dermatology, and neurology was an example of administrative change for the dean's office and its work group to consider. In addition, the departments of orthopedic surgery, physical medicine and rehabilitation, and urologic surgery also began to share administrative functions. These pilot collaborations eased the official introduction of administrative centers within the medical school. As discussed in Greenhalgh et al,2 “Innovations with which the intended users can experiment on a limited basis are adopted and assimilated more easily.”
Revisiting the idea of department clusters
In its 1998 report, the work group recommended administrative clustering of departments throughout the medical school. The dean and the director of finance followed up with reluctant department heads, suggesting, among other things, that they talk to the departments that were already sharing services to find out about the benefits of clustering. After a several-month campaign that included letters, one-on-one appeals, and many meetings, only four department heads remained resistant to departmental clustering. The dean then decided that no outliers would be allowed in the clinical departments and called for building five or six administrative centers with two to four departments clustered under each one.
The decision to create five or six administrative centers was based on an assessment of the factors that influence administrative workloads. For instance, the number of clinical faculty totaled about 650, with one department having more than 100 faculty members and, as a result, a greater administrative workload. The amount of sponsored research within departments totaled $97 million, with two individual departments responsible for about $16 million each—another contributing factor to administrative workloads. In light of this imbalance of departmental administrative workloads, deciding which departments to cluster together was a daunting task; in the end, it took two attempts. The first attempt was a blind approach that assigned numeric values to department attributes, such as number of faculty, support staff, and divisions, as well as amounts of space and sponsored and nonsponsored expenditures. Departments were grouped so that the total numeric values of each cluster were more or less equal. The resulting clusters did not make sense, however, either culturally or programmatically. For example, one cluster included the departments of anesthesiology, obstetrics-gynecology, ophthalmology, psychiatry, and radiology—departments whose numeric attributes added up correctly but that had few programmatic connections.
The second attempt was much more subjective. First, it was decided to maintain the two clusters (medicine, dermatology, and neurology; and orthopedic surgery, physical medicine and rehabilitation, and urologic surgery) that were already operating successfully. Then, all hospital-based departments were clustered together. The final decisions about which remaining departments belonged in which clusters were made with consideration for how specialties might mesh—and also with knowledge of department heads' personalities. In hindsight, the smoothest transitions were accomplished in administrative centers that included an “anchor” department, one that already had a robust administrative unit on which the center could build to provide services to smaller or midsize departments. Initially, six administrative centers were created; they eventually coalesced into five.
Once the administrative centers were established, the dean's office recognized an opportunity to extend the advantages of shared administrative services to the medical school's research institutes, multimillion dollar enterprises such as the Stem Cell Institute, the Biomedical Engineering Institute, and the Institute for Human Genetics.
Implementing the change
There were four key objectives in developing the administrative center model:
- Departmental access to high-quality administrative services at a reasonable cost
- Improved oversight of administrative functions within departments
- Enhanced recruitment and retention of professional administrative expertise
- Increased opportunities for developing best practices, policies, and standards
As of late 2005, the administrative centers had evolved and encompassed the following groupings:
- ▪ALRT—departments of anesthesiology, laboratory medicine, radiology, and therapeutic radiology, as well as the Institute of Human Genetics, the Biomedical Engineering Institute, and the Center for Magnetic Resonance Research
- ▪MEND—departments of medicine, emergency medicine, neurology, and dermatology, as well as the Stem Cell Institute
- ▪OPEN—departments of obstetrics-gynecology, pediatrics, otolaryngology, and neurosurgery
- ▪SOUP—departments of surgery, orthopedics, urology, and physical medicine and rehabilitation, as well as the Lillehei Heart Institute
- ▪FM/Ophthal/Psych—family medicine, ophthalmology, and psychiatry
For each cluster of clinical departments and research centers, each administrative center provides services in the areas of finance, human resources, research support, and education support. The administrative centers have differentiated to meet the needs of their departments, so that today, for example, the SOUP administrative center differs somewhat in structure and in culture from the ALRT center. We have illustrated the specific structure of the ALRT center as an example in Figure 1. Some functions, such as graduate medical education (GME), remain primarily based in departments. To deliver their services, some administrative centers place second-tier administrators throughout their departments, down to the division level. Others deliver those services directly, through center staff. All administrative center personnel, however, share best practices and work together to help the departments and the medical school achieve their goals.
Each administrative center is headed by an experienced, service-oriented director who manages the needs of his or her clinical department heads and faculty members. Figure 2 illustrates a center director's professional relationships within the center and the university. This position requires a mature person with several years of management experience in an academic department, an academic medical center, a health care organization, a faculty practice plan, or another complex service organization. Along with well-developed management abilities, the director in such a highly responsible position needs sound judgment, solid communication abilities, and high levels of interpersonal skills. Fortunately, the medical school was able to secure an effective blend of internal incumbents and external recruits to staff these demanding positions.
Evolution and Experience
Since instituting a new system of administrative centers at the medical school in late 1998, the dean's office, the department heads, the new center directors, and the faculty have responded creatively to ongoing challenges to culture, roles, and individuals.
Within six months of the change, the most vocal opponent among the department heads, who initially said the department would not join an administrative center, had changed his mind. He told the medical school dean's office that clustering departments under administrative centers was the best thing the school had ever done for the department.
The smaller clinical departments, in particular, realized the resources they had not been receiving as independent entities. Whereas larger departments had employed several administrative staff, smaller departments had to divide administrative tasks among fewer people. In smaller departments, single administrators were expected to be proficient in all functional areas and were unable to meet such demands. It was simply not possible for one person to be the department's chief financial officer, human resources director, information technology guru, facilities manager, and so on.
Under the new system, even the small departments would have access to top-notch expertise, which would otherwise be prohibitively expensive. Having an anchor department with robust resources, noted one center director, is an important support for this administrative model. Wisely grouping departments also helps, this center director said, because they might share the same culture, the same approaches to faculty compensation, or other similarities that ease the clustering process.
Despite the shared administrative responsibilities of the new model, some important departmental distinctions were retained. This new model increased dean's office oversight of departments, but administrative centers have not consolidated departments and have not taken away authority from department heads. Department heads are essential to the administrative structure, noted one administrative center director; their support is necessary for the administrative centers to work. Although it was a difficult transition for some department heads, who no longer had their own staff members reporting exclusively to them, they recognized that the new system presents fewer risks, provides consistent and high-quality service, and offers more transparency.
As a result of clustering, however, center directors not only report to the dean's office; they also must balance the needs and demands of three to four different department heads. “As a center director, there's never enough of you to go around for everybody,” said one. “[Yet] we provide better service than they ever received from one administrator” (L. Kenny, personal communication, 2005). One center director noted that the nature of this balancing act not only depends on the size of the departments involved but also on the character of the department heads. Some department heads prefer regular face-to-face meetings with administrative center directors, whereas others are content with frequent e-mail communications about progress with their projects. One administrative center director invites department heads to the center's monthly all-staff meeting. The center director said: “We try to connect the work that we do with their mission” (L. Kenny, personal communication, 2005).
Administrative center directors also face the challenge of justifying their budgets to department heads, who now, after working closely with center directors, may be more aware of these administrative costs than in the past. To help the department heads understand the value of the administrative centers, administrative directors offer them in-depth analyses of the department's finances, including how the clinical practices are performing financially. In-depth orientations for new faculty, including the culture of the clinical, educational, and research enterprises at the medical school, also improve the faculty's perceptions of the value of the centers' work.
Infrequently, conflicts between department heads and administrative center directors have arisen over such issues as the financial viability of departmental plans. When this occurs, personnel in the dean's office have been able to act as mediators to resolve the conflict.
Understanding how to staff the centers was difficult, particularly in the transition immediately after the launch of this new model. One administrative center director noted that she delegated too much work to a very competent employee, which caused that employee to experience burnout. Another director addressed the challenge of bringing together staffers from formerly separate departments into a new team. “It's like changing someone's personality” (L. Kenny, personal communication, 2005). She used regular newsletters, team-building volunteer activities outside the workplace, and other communication tools to establish connections among new coworkers, but she said that it took years to achieve identity as an administrative center and that she still strives to improve teamwork.
After several years of experience, the center directors have better knowledge of how to help staff members succeed, how to train them in cross coverage, and how to retain the talented people who best serve department heads, faculty, and the medical school's leadership. Because the role of center director is complex and must be filled by a seasoned administrator, the job of administrative center director is perhaps the most challenging one to fill in this new model. Recently, the medical school had two failed attempts before successfully filling an opening for a new director for the MEND center.
Since the administrative centers were launched, the university has instituted several new information systems as well as numerous new processes and procedures related to grants management, budget development and management, and GME. In addition, as in all research institutions, the University of Minnesota must manage increased demands from the regulatory sector. Yet, the system of administrative centers has demonstrated resilience and flexibility in this changing environment. Designed to be centralized enough to promote efficiency, yet decentralized enough to encourage understanding and timely responses to local problems, the administrative centers represent an effective management approach within this complex enterprise.
The medical school completed an internal progress report on the administrative centers in 2001. That report noted that the initial groupings of the administrative centers had been reshuffled and that the basic science departments remained outside this model. That was because those departments were being reorganized in cooperation with the college of biological sciences, but the medical school dean at the time intended that they be incorporated into the model of administrative centers in the future. In addition, integration of medical school administrative centers with the University of Minnesota Physicians' clinical service units was an ongoing issue to be addressed. Finally, the administrative center directors needed more support and guidance from the dean's office to staff the centers and to apply best practices consistently. We fully addressed these issues when an external evaluation was conducted.
Medical school dean Deborah Powell employed ECG Management Consultants to do an independent evaluation of the administrative center model in 2005. The confidential, in-depth evaluation that resulted included a global assessment of the model. The consultants found that, as currently configured, the finance and human resources functions of departments require the focused expertise and use of information systems that are best provided at the administrative center level.
Although the medical school has a designated institutional official for GME in the dean's office, there remain GME responsibilities that are highly dependent on the attributes and activities of the departments within each center. Thus, the current, primarily department-based model for these GME functions is appropriate. Based on discussions with medical school GME leadership, however, it is recommended that an administrative center-level GME supervisor role be instituted to provide consistency among departments. This position could be filled by an existing staff member.
The consultants also noted that responsibilities and reporting roles were not always consistent across centers; the medical school director of finance and administration would officially take on administrative oversight of the centers to address these inconsistencies. They also suggested that the chief medical officer of the faculty practice plan be primarily responsible for managing center director roles in the clinical practice arena. Finally, the report also included the results of a survey of department heads that revealed substantial support for the administrative center model.
The ECG consultants noted that, as organized in 2005, the administrative center model was saving the medical school a considerable sum of money compared with stand-alone departmental administrative structures. They came to this conclusion by comparing the costs of the medical school's administrative center model with those of stand-alone large, midsize, and small departments at the many medical schools where ECG has consulted. For understandable privacy reasons, they prefer that these cost comparisons remain confidential.
The report also noted that, given the growth in faculty and increased demands for regulatory compliance, the center directors might be overextended. The report suggested adding at least one administrative center to lighten the administrative burden on the current center directors and provided a few alternative groupings of departments under administrative centers, noting the associated incremental costs.
Effective July 2006, a sixth clinical administrative center was formed, and, in 2007, the basic science departments will be brought into this administrative structure.
A Structure of Strength and Flexibility
There are about 1,300 full-time faculty at the medical school, and more than 1,100 of them are clinical faculty. Although developing administrative centers under which departments are clustered might seem a simple change, the results have been impressive.
Administration processes in the medical school have been dramatically transformed and streamlined in the last decade. The previously fragmented administration system has been integrated into six administrative centers that function with more professional expertise and with more effectiveness than did the individual departments on their own. Leaving department heads in charge of day-to-day departmental operations, the dean's office gains an important measure of centralization through the evolving administrative structure, which may promote understanding of the interdependency of the whole medical school and certainly aids in reducing redundancy of staffing.3 As a result, this streamlined system saves money for the school—up to $3 million annually, according to the recent evaluation by ECG Management Consultants.
In the seven years since the administrative center model was implemented throughout the medical school, the center directors have addressed many issues, including appropriate staffing and workloads for employees, career development for employees with the goal of employee retention, and, most importantly, changing the administrative culture to be responsive to department heads and faculty while also reporting to the dean's office. The administrative structure at the medical school continues to evolve as the faculty grows and departmental needs change. We believe, however, that we have developed a structure of strength and flexibility that can respond to our school's issues today and tomorrow.
Contributors include Marcie Christensen, Linda Kenny, Dolly Schmidt, Jon Pryor, MD, and Deborah Powell, MD, dean of the medical school. Research help was provided by Kathleen Watson, MD, medical school associate dean for students and student learning, and Jim Begun, PhD, professor of health care management, School of Public Health.