Several well-documented trends are shaping the delivery of ambulatory care. Health care costs have escalated each year for the last two decades.1,2 Physicians find that they have less time to spend addressing each patient's clinical issues,3,4 while they must spend more time entering patient data into their computers.5 It is not surprising that some physicians are questioning their ability to do what they were trained to do, and some are leaving or considering leaving the practice of medicine, particularly in primary care internal medicine.6–8 A report by the Massachusetts Medical Society9 documenting acute shortages in certain specialties and a growing shortage of primary care physicians within the state, called for changes in policy to stop these trends. The critical need to retain primary care physicians demands that attention be paid to the quality of the work environment in which they practice, including the support they receive from nurses, medical assistants, and support staff.
At the same time, the families of health care workers are also encountering difficulties. Starting in the 1970s, as it became more and more difficult for families to make ends meet on one income, women joined the workforce in increasing numbers.10 Additionally, educational opportunities for women expanded, and previously male-dominated professions—like medicine—began to train women, so that in the United States, 48% of all medical school graduates are now women,11 and 28% of all practicing physicians are women.12 Although expanded educational and employment opportunities for women are positive, the workplace has not been restructured to support both women and men in fulfilling their roles at work and at home.
The changes in health care delivery and the changes in family life pose a dual challenge to health care institutions providing ambulatory care. In response, we need a dual goal to both (1) increase the effectiveness and quality of primary care and medical specialties and (2) provide a satisfying work environment that allows health care providers—including physicians, physicians' assistants, nurses, medical assistants, and support staff—to meet their responsibilities to their patients and their own families. This dual goal—enhanced patient care and an improved work environment for employees—is what we call the “dual agenda,” and it is the basic premise that underlies our work.13 The question is, What changes will be required in our health care institutions to achieve this dual agenda?
In May 2005, the Gretchen and Edward Fish Center for Women's Health, a small ambulatory practice offering primary care and gynecological care to women at an academic medical center, relocated to a suburban community one mile from the main hospital campus. By increasing the number of primary care physicians, recruiting physicians in 10 other medical specialties, and encouraging cross-specialty collaboration, the center began providing a holistic approach to multispecialty ambulatory care for women.
However, translating a new concept for women's medical care into a viable organizational reality posed new and complex issues for the work design of the practice. Collaborating with a team from the Sloan School of Management at the Massachusetts Institute of Technology (MIT), the center's leaders began a pilot project with great potential to transform both the health care experience of patients and the work experience of providers both on and off the job. The goal was to build a “quality-of-life learning laboratory” to complement the already-established “clinical learning laboratory.”
The new practice was designed with four specialized clinical areas: a primary care area, a gynecology/urogynecology area, an area for dermatology, nutrition, and mental health services, and a medical specialties area including endocrinology and diabetes treatment, cardiology, rheumatology, physiatry, gastroenterology, and thoracic surgery that focused on women at risk for or diagnosed with lung cancer. Dermatology, nutrition, and mental health were combined as a team because of the space needs of these three specialties and the desirability of a second, more private waiting area for mental health. Whereas the clinical work was subdivided by specialty, the administrative support and the registered nurse triage call-in lines were centralized, both spatially and functionally.
Once fully staffed, the center employed 35 physicians and accommodated 35,000 patient visits per year. The demographic and occupational characteristics of the workforce during the time of the pilot project are described in Table 1.
Data collection and work redesign process
The pilot project was designed using collaborative interactive action research,14 a method developed in private-sector workplaces. The center/MIT team adapted it to the specific conditions, particularly the time-squeezed schedules, of the health care service environment. This method is based on the assumption that workplace structures and practices are not immutable but, rather, can be changed to accommodate a dual agenda of increased work effectiveness and enhanced work/life satisfaction. It is defined by inclusive participation of all staff in deciding what changes to make and by transparency regarding proposals and results. The concept of fluid expertise provides a foundation of mutual respect, affirming that both outsiders (MIT) and insiders (all center staff) hold knowledge that is needed to make the change process a success.15
The collaboration unfolded through the establishment of a liaison committee comprising a cross-section of practice staff, management, and a three-person research team from MIT. Together, through regular meetings, they developed a four-step process for change.
Step 1: Assessing the workplace.
The MIT team, using confidential, open-ended interviews, gathered information from 60 staff members about their experiences, both on the job and in their personal/family lives. In addition, the MIT team observed 100 hours during clinical practice to record patient–provider and provider–provider interactions. Because internal review board approval for this study precluded any observation inside patient exam rooms, observations of patient–provider interactions were limited to waiting areas and practice spaces adjacent to exam rooms where data on vital signs are taken and recorded.
The MIT team synthesized those findings and presented them to the staff at meetings in each of the four clinical areas, communicating both what was working well and what was not working well at the center. Staff largely affirmed, and in some cases elaborated, these findings. The key problems identified included
* top-down decision making and employees' lack of voice in key issues affecting the practice, resulting in a lack of cohesion and community;
* centralization of administrative and triage functions not organized to take into account the particular needs of each specialty and the four clinical areas;
* insufficient coverage and cross-training;
* difficulties managing responsibilities for families and other personal needs while working; and
* limited flexible work options that fit the needs of some, but not all, workers or all clinical operations.
Step 2: Redesigning the work.
Having agreed on problem areas in the practice, staff were asked to generate ideas for new ways of organizing the work. The building blocks of this work redesign process were the specific knowledge of staff in all occupations who understand the nature of problems in their daily work with patients.
The key features of the pilot project reflect the key problem areas identified in the initial workplace assessment.
* Care teams. To address employees' lack of voice in decision making, cross-occupational care teams were established. The idea was to create a regular forum that would bring together physicians, medical staff (which included nurses and medical assistants), and administrative staff in each clinical area to identify problems in the delivery of care and empower them to come up with solutions for improving the patient experience and their own daily work experience.
* Decentralization. Under the “centralized model,” administrative support staff were trained to work in any of the practice's 12 specialties. This provided flexibility to place them in different clinical areas as needed, but it did not allow them to develop a deeper understanding of particular specialties, which was identified as a weakness in the practice. To remedy this problem, a “decentralization” plan was devised in which all support staff would be trained to work with primary care patients but would also develop secondary areas of expertise in one or two clinical areas, linking their administrative and clinical knowledge and developing a deeper fluency. Support staff were divided among gynecology, medical specialties, primary care, and dermatology/mental health/nutrition, and became members of the care teams linked to those areas.
* Cross-training. To become proficient in new areas of specialization, and to equip the practice to be nimble and responsive to staff absences and leaves, additional training for support staff was provided so that they could develop a deeper knowledge base to meet the needs of patients.
* Team approach to flexibility. To address the difficulty of integrating work and family responsibilities and to find a more effective and equitable way of dispensing flexible work options, staff were encouraged to discuss and respond to requests in their teams, rather than handle them through individual negotiations with the practice manager as they had done in the past.
Step 3: Implementing the pilot project.
The implementation of the pilot project was carried out across six months, and not all components were fully realized. The care teams were established in January 2007, although meeting frequency and attendance varied by clinical areas. The scheduling system and phone system were reconfigured in early 2007 to accommodate the decentralized model of work organization, and specialty-specific training was carried out across six months, with additional training opportunities as needed.
Step 4: Evaluating the pilot project.
The pilot project was evaluated through questionnaires that queried staff about their experience on their teams and in their clinical work. Three were administered as paper surveys, and one was Web based. Staff were asked to provide answers “on the basis of your observations and experiences last week,” with the dates specified. The final, Web-based survey had additional questions to gauge staff's attitudes regarding their own care teams. Finally, some quantitative data were obtained on staff turnover rates and phone records that tracked patient wait times and abandoned patient calls.
Table 1 shows the demographics of the 60 staff members. All but two study participants were women, and although the majority were white, there was significant racial and ethnic diversity (30%). A range of occupations were represented.
All four surveys had a response rate of about 55%. The aggregated findings on questions related to individual work experiences were inconclusive and/or mixed in terms of showing positive or negative impact in the operations of the practice. We believe two things may explain these results: (a) we provided too short a period between surveys to surface consistent change in operational practices, and (b) responses were heavily impacted by specific clinician or staff absences and/or significant patient events in that particular week.
Table 2 displays the staff's attitudes toward their own care teams. The responses provide early evidence of the positive impact of the newly established cross-occupational care teams on the daily experiences of staff and the use of the teams as a forum for collective problem solving in three out of four clinical areas of the practice.
The overriding shift during the pilot project was toward a new approach to quality improvement that focused on the needs of the workforce and the culture, structures, and practices of the workplace. Before the pilot project, problem solving and quality improvement focused mainly on improving the patient experience. By introducing the dual agenda, the pilot project applied improvement efforts to both the workforce experience and the patient experience.
The process of collaborative work redesign
The process of redesigning work structures was just as important as the new practices and organizational forms that emerged. The process was, in essence, a “training” in new ways of working together in the practice. There were three key components of this learning process.
By participating in the collaborative interactive action research process, physicians and staff learned how to assess the daily operations in their own clinical areas. Through meetings with the MIT team, they learned to identify what worked well and what constituted effective practice, as well as what was not working well and how to analyze why.
Sharing their own observations and knowledge to analyze problems and generate solutions empowered staff at all levels. This often created a new level of collaboration which enhanced coordination across occupations, increased cooperation between staff in different clinical areas that shared common administrative staff, and fostered an exchange about workplace problems that was new and exciting for all involved. This relational coordination has been shown to have positive effects for patients in other health care venues, notably nursing homes.16
Part of the learning was the development of new models of leadership. The common hierarchy in medicine in which the physician is seen as the bearer of all essential wisdom, the single voice of authority, was challenged. It was replaced with a model in which staff at all levels, including administrative and clinical support staff, are valued and seen as potential leaders for improving clinical care and enhancing patient satisfaction.
A team approach
The most significant success of the work redesign experiment was the establishment of self-managed, cross-occupational care teams in three out of the center's four clinical areas. Each team was given the choice of determining how frequently it would meet, how agendas would be set, and who would lead the team. Several observable changes took place.
Being part of the solution.
During the initial data collection process in 2005, many physicians and staff used the individual interviews to articulate grievances about the center and its management. Once the teams were formed, staff took increased responsibility for addressing issues in their area. This signaled a cultural shift in attitude from complaining to solving problems, and this process was done collectively and publicly, not one-to-one and privately.
Staff were able to engage in effective problem solving by creating meaningful dialogue across occupations. There are examples of support staff telling physicians that they needed better information and support to do their jobs effectively, and examples of physicians offering to change their ways of operating to be less individualized and more team-oriented. In one team, the coleaders were a physician and a support staff person who together spearheaded new systems for their area.
Employee voice enhanced.
Whereas physicians had some input into practice operations before the pilot project, other staff had virtually none. The creation and development of the teams meant real involvement in practice issues for staff at all levels, including administrative support staff who previously met only with other support staff and the practice coordinator, never with the physicians and other professional staff.
These changes are documented quantitatively in the high levels of attendance recorded for most team meetings and in responses to survey questions about whether the teams had improved matters, maintained the status quo, or made things worse (see Table 2). The scores, although somewhat variable by clinical area, were largely positive, and the most highly functioning teams had the highest satisfaction scores. For example, members of the gynecology care team, which was co-led by a physician and a staff member, said that team meetings were very productive (4–5 range on a 5-point scale where 1 = not productive and 5 = highly productive), and they reported being able to bring up their concerns in meetings (4–5 range on a 5-point scale where 1 = not at all and 5 = completely). As one medical assistant wrote on her survey, “We get to hear and talk about the things that are working or not working for us and how or what we can do to make things better.” The medical specialties group, which, as mentioned before, consists of individual physicians combined into one group with shared administrative and clinical support staff, were, because of their clinical schedules, seldom present in the clinic at the same time and therefore had difficulty meeting as a care team. However, the new ways of working within the center affected their work and attitudes toward work, mainly through their close interaction with the gynecology care team.
Flexible work arrangements
Before the pilot project started, flexible work options were available, but they were neither equitably distributed nor effective for the delivery of quality care. Part-time work options were available for physicians, nurses, and other professionals, but not for medical assistants and administrative staff. Flexible hours were available to full-time workers to allow for family responsibilities or to attend school after work, but leaving early or coming in late often created problems for coworkers and patients.
Away from individual accommodation to a team approach.
The teams became the venue for discussing how to organize flexible work options, replacing the former method of the manager addressing the needs of individual workers who needed to come in late, leave early, or take time off. There was increased understanding that all staff, regardless of occupation, have needs for flexibility and families who depend on them, and a growing ability to meet individual needs in ways that work for all team members and for patients. For example, a staff member returning from maternity leave discussed with her team that she would like to change her work hours to accommodate her child's day care situation. The team discussed and jointly agreed to changes in their schedules that provided staff coverage to support patient and physician needs that also met the needs of individual team members.
New training opportunities for all staff increased their ability to do their jobs with the most up-to-date clinical and operational information. It also increased the ability of nurses, medical assistants, and administrative staff to work outside their own clinical area in case of others' absences, late arrivals, or early departures. This willingness to “float” and learn the procedures of a specialty outside of one's regular job increased the effectiveness of the clinical operations and the satisfaction that the support staff have in their work.
Discussion and Conclusions
As the health care environment rapidly changes, a successful ambulatory medical practice is one in which senior managers, physicians, and all other staff can publicly identify and discuss problems, think through collective solutions, and effectively and collaboratively embrace change. The leadership at the center has embraced a continuing process of change, and, through the development of cross-occupational care teams, physicians and staff have been given both a forum for discussing change and the tools to think about new ways of working together. This type of collaborative work will only become more important as ambulatory medical practices are expected to improve patient outcomes while streamlining operations.
Several aspects of the pilot project developed at this center suggest a model for innovation and change appropriate to other ambulatory practices. First, the center's physician and administrative leadership is allowing—in fact, nurturing—ongoing experimentation in work organization and work culture. Although changes in work organization are sometimes included in health care quality improvement efforts, analysis and experimentation by frontline workers is often overlooked, even though it is necessary to creating long-term, sustainable change.
Linked to this focus on work organization is a related focus on the workforce itself. By embracing the concept of the dual agenda, the center is creating innovations in work design and work culture that enhance not only the quality of clinical care for patients but also the quality of life for the caregivers. There is also recognition of the importance of developing leadership among faculty and staff, that members of the practice workforce must be given not only appropriate clinical training in primary care and other medical specialties but also the skills to lead an ongoing process of quality improvement in their clinical areas. When all members of the practice are seen as potential leaders, issues of cost containment, clinical effectiveness, and patient satisfaction become challenges not only for management but for the entire workforce.
Second, the center's vision of delivering multidisciplinary, gender-specific care is a model that can enrich the day-to-day practice environment for physicians by providing opportunities for clinical and research collaborations between primary care physicians and the medical specialists. Creating time for continuing professional development—especially when increasing administrative burdens threaten physician work satisfaction—may help the recruitment and retention of physicians in the areas of internal medicine and other specialties where shortages are already evident. Although the center's focus is gender-specific, collaborative interactive action research and the cross-occupational team model can apply to many practice types, both within and outside of academic medicine.
Third, this pilot program created a climate in which all physicians and staff expected the right to satisfying lives at work, at home, and in their communities. To meet this expectation, the center adopted a model of flexible work arrangements that meshed both provider and patient needs. This explicit acceptance of the roles and responsibilities that health care providers have in their private lives outside of the workplace is key to increasing morale and decreasing absenteeism and turnover. A better-supported and less stressed health care workforce will ultimately benefit both patients and providers.
Fourth, we believe that this pilot study has important implications for academic medical centers as places of learning and innovation. Team-based care is an ever more common theme in practice to achieve better clinical outcomes and patient satisfaction,17 and it requires that medical education include training on how to build and sustain effective teams. We hypothesize that these cross-occupational care teams can serve as important “learning communities” in which to improve the quality of patient care, streamline operations of a practice and/or medical service, and address issues of the quality of work life, including integration with other aspects of one's family/personal life outside of the work setting. Challenging the traditional hierarchy of academic medicine is a hurdle that will have to be surmounted if we are to truly adopt a team-based model that addresses all aspects of medical practice.
Our study has limitations that will need to be considered as collaborative, interactive action research and cross-occupational teams are adapted to other practice settings. We undertook this pilot project in a single practice site where the mission is to provide gender-specific health care integrated across specialties and disciplines. Both the workforce and the patient population were predominantly female. One might hypothesize that women may be more likely to challenge the traditional hierarchy and more likely to develop a model of distributed leadership, though in the center's experience—and in medicine generally—there is no evidence to support this hypothesis. We had neither a practice with a different clinical focus nor a similarly focused practice that had not undertaken a collaborative, interactive action research intervention with which to compare our process and outcomes. Another limitation, due to the study design and IRB restrictions, was that we were not able to collect patient-specific data to evaluate patient outcomes before and after the intervention. We were also unable to compare patient satisfaction data because of a change in the institution's ambulatory patient satisfaction methodology from an in-house tool to a nationally recognized external survey during the course of the pilot project. Lastly, there was a considerable commitment of time by the faculty and administrative leadership, as well as the staff, to attend meetings, and not all ambulatory care practices can spend the time needed for this kind of process. The MIT team also invested time preparing for meetings and debriefing among themselves and with the faculty and administrative leadership of the center, especially in the first year of the study. As with any new team undertaking a change process, an initial investment of time is necessary to engage, build trust, and jointly commit to learning the skills needed for new ways of working.
The cross-occupational teams have continued to work on quality and process improvement as well as work–life integration in the year since the end of the pilot project. But once a workplace change process has begun, it is important to keep asking questions. What obstacles may prevent further change? What action steps should be taken next? In naming obstacles and sustaining innovation, it is important to remember that the best solutions often come from those most directly familiar with the work—the cross-occupational care teams. Giving physicians and staff a voice and a substantive role in creating new work systems and practices has had a positive impact on the center. Structural and cultural change can go hand in hand, and sometimes the cultural change process is the hardest of all. The idea that responsibility for change must be shared and that solutions to individual needs must be addressed collectively is a huge shift in workplace culture, and it may be the most significant change that has taken place at the Fish Center for Women's Health.
The authors wish to thank the faculty and staff at the Fish Center for Women's Health at Brigham and Women's Hospital for their participation and support.
This study was supported by the Alfred P. Sloan Foundation.
IRB approval was obtained for this study from the Partners Human Research Committee, Partners HealthCare, and the Committee on the Use of Humans as Experimental Subjects at MIT.