Improvement Projects Led by Unit-Based Teams of Nurse, Physician, and Quality Leaders Reduce Infections, Lower Costs, Improve Patient Satisfaction, and Nurse–Physician Communication
Improvement Projects Led by Unit-Based Teams of Nurse, Physician, and Quality Leaders Reduce Infections, Lower Costs, Improve Patient Satisfaction, and Nurse–Physician Communication
The University of Pennsylvania Health System uses unit-based clinical leadership teams, composed of a physician leader, nurse leader, and quality/safety project manager, to develop unit-specific initiatives designed to improve quality and patient safety. Team members meet formally and informally to develop and implement initiatives, with sharing of best practices across teams. Working collaboratively allows the nurse, physician, and quality leaders to develop closer working relationships and to create a sense of shared accountability for unit performance, a culture that spreads to other clinical staff as they implement these initiatives. The program reduced central line–associated bloodstream infections (and the cost of treating such infections), urinary tract infections and pressure ulcers, improved adherence to health system standard for medication reconciliation, increased reporting of errors and near-misses, and led to higher patient satisfaction and improved nurse–physician communication and teamwork.
Poor nurse–physician communication is common and frequently leads to negative consequences for patients and health care organizations.Most hospitals and health systems do not implement formal initiatives to facilitate the development of strong working relationships between physicians and nurses.
- Poor nurse–physician communication: Nurses and physicians often work in silos, leading to poor communication. Surveys of nurses at the University of Pennsylvania Health System found that nurse–physician relationships, including the quality of communication and collaboration, remained suboptimal, with communication occurring only on a basic level and often without a full understanding of each others' roles, challenges, and work processes. Communication between nurses and physicians (including rotating residents and medical students) at this busy academic medical center typically occurred only informally, either through daily patient care or via written medical record documentation. Furthermore, the health system's unit-level initiatives—whether geared toward improvement of quality, safety, or patient satisfaction—tended to be driven by nurses, with physicians often not fully or consistently engaged.
- With negative consequences: Ineffective nurse–physician communication can lead to poor patient outcomes and can hinder implementation of initiatives and guidelines to improve outcomes.
- Failure to support good communication and collaboration: Most organizations do not have formal strategies for ensuring that nurses and physicians communicate well and collaborate on ongoing initiatives to improve quality.
Description of the Innovative Activity
The University of Pennsylvania Health System uses unit-based clinical leadership teams, composed of a physician leader, nurse leader, and quality/safety project manager, to develop unit-specific initiatives designed to improve quality, patient safety, and satisfaction. Team members meet formally and informally to develop and implement initiatives, with sharing of best practices across teams. Implementation of these initiatives typically involves other clinical staff, who then work together to improve the targeted outcomes. Key elements of the program include the following:
- Egalitarian-based, unit-focused teams: Each team consists of a physician leader, a nurse leader, and a quality/safety project manager, although teams often ask other individuals to participate or advise, as needed. Although most teams are dedicated to a specific unit, some involve more than one unit, with 18 teams being responsible for 22 units (including five intensive care units and women's health services) as of June 2011 (expanded from 5 original teams at program inception). For example, one team covers two general medicine units that share a nurse manager, while another team covers three oncology units. Some teams also have a slightly different composition—for example, a surgical unit team has two physician coleaders who represent the two different types of surgical patients treated on the unit. Four full-time quality/safety project managers divide their time among the 18 teams. The teams use an egalitarian model in which each member has an equal leadership stake and shares accountability for results. The key roles of team members are outlined below:
- Physician leader: The physician leader brings the physician perspective to the development of initiatives, and communicates the importance of the projects to physician colleagues who admit to the unit or whose teams care for patients on the unit.
- Nurse manager: The nurse manager helps to implement the initiatives in a realistic manner, taking into consideration the perspectives of bedside nurses, and communicates initiative specifications to frontline staff.
- Quality/safety project manager: The quality/safety project managers gather unit-specific data to inform the development of initiatives, help translate the data into actionable initiatives at the bedside, and offer practical assistance with project management. They monitor patient feedback and provide concurrent review of safety incident reports for their assigned units. They periodically participate in interdisciplinary unit rounds (with nurses and physicians) to observe care processes, and also monitor activities across all unit-based leadership teams. When similar themes arise, the quality leaders link teams to enable them to share best practices, allowing teams to adopt or adapt initiatives without “reinventing the wheel.” (See “Sharing of best practices across teams” below for more details.)
- Other team members: Information provided in June 2011 indicates that additional individuals participating on the teams include physician specialists and unit-based clinical pharmacists (pharmacists assigned to each unit).
- Project selection based on organizational priorities: The nurse manager, physician leader, and quality/safety project manager develop initiatives centered around the Blueprint for Quality, a health system document that outlines clinical imperatives for the coming year (such as transitions in care, reducing variations in care, coordination of care, and shared accountability). Each team selects areas of focus based on the needs of the particular unit. However, several common priorities have emerged, including initiatives aimed at reducing catheter-related bloodstream infections, reducing urinary tract infections, improving patient satisfaction, and improving transitions in care so as to reduce unplanned readmissions. Units that tackle similar problems often vary in their approaches; for example, reducing bloodstream infections on an oncology unit requires a different strategy than on a cardiac surgery unit.
- Initiative development process: The team uses a combination of formal meetings and informal communication (via e-mail and telephone) to review data and develop and implement initiatives, as outlined below:
- Weekly meeting to review data, develop action plans: The quality/safety project manager brings unit-specific data (e.g., incidence of deep vein thrombosis, central line–associated bloodstream infections, and urinary tract infections; incident reports) to a weekly, unit-based, clinical leadership team meeting. (With the assistance of health system information technology staff, program leaders are currently developing a data warehouse to facilitate real-time data provision to the unit-based teams.) The team uses the data to identify and discuss potential opportunities for improvement, and then develops an action plan in select areas. The nurse and physician leader implement the plan by working with their colleagues.
- Ongoing informal communication: The team engages in an ongoing dialogue via telephone and e-mail, as needed, to discuss project development, patient safety, incident reports, and other unit-related issues.
- Monthly meeting with other teams, system executives: Each month, one of the weekly meetings includes groupings of unit-based clinical leaders and the health system chief medical officer and chief nursing executive. At these sessions, the teams present a status report on ongoing projects. Having high-level system leaders at this session gives the teams a chance to request advice and make the case for additional resources, if needed.
- Implementation by frontline physicians and nurses: Physician, nurse, and quality leaders describe the planned initiatives to key clinical teams who work on the unit; these staff then work together to implement the initiatives during daily patient care. This collaboration provides an opportunity for nurses, physicians, and other clinicians to learn to communicate more effectively with each other, both about the initiative and about patient care more generally.
- Sharing of best practices across teams: Teams share best practices through the following mechanisms:
- Quality/safety project managers: As noted above, four quality/safety project managers work across multiple teams, thus facilitating cross-team communication about similar initiatives.
- Interhospital quality exchange: Program developers have initiated “quality exchanges” between the Hospital of the University of Pennsylvania, Pennsylvania Hospital, and Penn Presbyterian Hospital. For example, the Hospital of the University of Pennsylvania teams that successfully reduced bloodstream infections hosted teams from Pennsylvania Hospital; in turn, Pennsylvania Hospital teams that reduced ventilator-associated pneumonia hosted teams from Hospital of the University of Pennsylvania. A third exchange to include all three hospitals is planned around the topic of reducing urinary tract infections.
- Chief nursing executive/chief medical officer council: Chief nursing executives and chief medical officers from the three health system hospitals, an affiliated rehabilitation facility, and a home care/hospice organization meet monthly to provide status updates and share best practices developed by teams within their own organizations.
- Quarterly rounding: Information provided in June 2011 indicates that each quarter, top hospital executives round with the unit-based clinical teams to learn about their activities. In addition, each quarter the hospital holds unit-based clinical team grand rounds to enable them to share information across clinical cohorts.
- Performance recognition program: Information provided in June 2011 indicates that the hospital has instituted a unit recognition program in which units are awarded a bronze plaque for going 500 days, a silver plaque for going 750 days, and a gold plaque for going at least 1,000 days without a hospital-acquired infection; plaques are awarded for days without a central line–associated bloodstream infection, ventilator-associated pneumonia case, and urinary tract infection. The vice president for quality and patient safety, the chief medical officer, the chief nurse executive, the chief operating officer, and other clinical and operational leaders attend a presentation to present the plaques and honor unit staff for their achievements.
Context of the Innovation
The University of Pennsylvania Health System, an academic health system that includes the Hospital of the University of Pennsylvania, Penn Presbyterian Medical Center, and Pennsylvania Hospital, has nearly 15,000 employees and handles roughly 80,000 admissions each year. Health system leaders strongly support efforts to improve quality and patient safety, and the Hospital of the University of Pennsylvania received Magnet designation from the American Nurses Credentialing Center in 2007. After the health system's chief medical officer and chief nursing executive began working together on quality initiatives designed to improve patient outcomes, care transitions, and patient satisfaction at the system level, they quickly realized that better physician–nurse partnerships at the unit level would be critical to improving both unit- and system-level performance. As noted earlier, internal surveys found that collaboration between nurses and physicians remained suboptimal, with little accountability for performance at the unit level.
The program reduced central line–associated bloodstream infections (and the cost of treating such infections), urinary tract infections, and pressure ulcers; improved adherence to health system standards related to medication reconciliation; increased reporting of errors and near-misses; and improved patient satisfaction and nurse and physician perceptions about the quality of communication and teamwork.
- Fewer central line–associated bloodstream infections (and lower infection-related costs): Thanks to a hospital-wide initiative, central line–associated bloodstream infections fell in all units at the Hospital of the University of Pennsylvania. However, between 2007 and 2008, the units that adopted the clinical leadership model experienced 33 fewer central line–associated bloodstream infections compared with similar units that did not use the model. Further analysis estimated a cost savings of $477,200 from this reduction, representing a net savings of $147,200 after accounting for program costs of $330,000. These estimates do not include the potential cost savings due to other quality improvement outcomes on participating units. Information provided in June 2011 indicates that the central line-associated bloodstream infection rate fell from 6.75 percent in the first quarter of fiscal year 2007 (pre-implementation) to 0.5 percent in the third quarter of fiscal year 2009 (post-implementation), a rate which has been sustained through the fourth quarter of fiscal year 2010. As measured on 13 units, days since the last central line–associated bloodstream infection ranged between 300 and 1,210 days as of May 31, 2011; four units have gone over 1,000 days without an infection.
- Fewer pressure ulcers: The percentage of patients experiencing pressure ulcers at Penn Presbyterian Medical Center decreased by more than 20 percent (from 1.9 to 1.5 percent) on the medical surgical units and by 15.9 percent (from 10.2 to 8.5 percent) on critical care units between fiscal year 2008 (pre-implementation) and fiscal year 2009 (post-implementation). Information provided in June 2011 indicates that the overall hospital-acquired pressure ulcer rate has fallen by 40 percent from pre-implementation levels.
- Decline in urinary tract infection rate: Information provided in June 2011 indicates that the hospital's urinary tract infection rate has fallen by 30 percent since implementation. As measured on 10 units, days since the last urinary tract infection ranged between 95 and 380 days as of May 31, 2011.
- Long time lag since last ventilator-acquired pneumonia: Information provided in June 2011 indicates that as measured on five units, days since the last ventilator-acquired pneumonia case ranged between 95 and 650 days as of May 31, 2011.
- Better adherence to medication reconciliation standards: Health system policy states that all patients should have their medications reconciled within 24 hours of hospital admission. Pre- and post-implementation comparisons at the Hospital of the University of Pennsylvania indicate that adherence to this standard increased by 44 percent (from 55.3 to 79.6 percent) in units that adopted the model, well above the 13.5-percent increase in nonadopting units (from 68.5 to 77.7 percent). Similar levels of improvement occurred at Penn Presbyterian Medical Center and Pennsylvania Hospital.
- More reporting of errors and near-misses: The number of reports submitted to the Hospital of the University of Pennsylvania's online reporting system grew at a faster rate on units that adopted the program than on those that did not; program developers believe that this is due to a greater culture of trust and communication on these units. (Hard data are unavailable.)
- Higher patient satisfaction: Patient satisfaction at Penn Presbyterian Medical Center increased between fiscal year 2008 (when the program was implemented) and fiscal year 2009, with increases in the following: staff response to patient pain (from 84.9 to 86.7 percent), staff teamwork in the provision of care (from 87.3 to 89.0 percent), and likelihood of recommending the hospital (from 66.9 to 70.0 percent).
- Clinician perceptions of better communication/teamwork: Anecdotal information and qualitative survey data from nurses and physicians working on units where the model has been implemented suggest that the program has improved communication and teamwork. Selected statements culled from surveys include the following:
- “Communication, interaction, and practice have improved as a result of the unit-based clinical leaders.”
- “I have definitely obtained a better understanding of things from the physician perspective and increased communication as much as possible.”
- “I think I always felt like I needed to do it all. It is nice to work with a group who have various interests and skills, and the unit benefits from all that expertise.”
- “I think the staff definitely feels more supported from a physician perspective. Our physician leader takes time to get to know the staff now, discusses issues with them, and looks for their feedback.”
- “Discharge rounds are much more efficient. The physician or nurse practitioner identifies patients who have additional needs.”
- “Orientation sessions have been enhanced to include more physician input and insight.”
Planning and Development Process
Key elements of the planning and development process included the following:
- Selection of physician leaders: Chairs of the various hospital departments/divisions suggested physicians who could serve as unit leaders on the teams. There is now an established appointment/reappointment process for the physician members of the unit-based clinical leadership teams.
- Presentations to unit staff: Quality/safety project managers attended nursing unit councils and staff meetings on targeted pilot units to discuss the initiative with bedside nurses.
- Initial pilot testing: The unit-based clinical leadership teams were piloted on five units. Although no formal training was provided, the chief medical officer and chief nursing executive explained the purpose of the program to team members, who were charged with developing a working relationship and pursuing two improvement projects related to any of four system objectives: transitions in care, reducing variations in care, coordination of care, and shared accountability. As part of the pilot, the teams held weekly meetings, implemented or refined interdisciplinary rounding programs, and oriented house staff to the needs of the unit and the role of the leadership team.
- Program expansion: Based on the success of the pilot test, by 2010 the program had expanded to include 13 teams covering 18 general care units at the Hospital of the University of Pennsylvania and is on all the units at Pennsylvania and Penn Presbyterian Hospitals. Information provided in June 2011 indicates that in 2011, the Hospital of the University of Pennsylvania expanded the program to additional units, including intensive care units and women's health services, and is planning further expansion to the emergency department and the transition care unit.
- Refining model: Following expansion in 2008, refinements were made in 2009 to improve the effectiveness of the teams as follows:
- Orientation program: Program leaders are developing a formal orientation program to train new nurse managers and physician leaders on the unit-based clinical leadership model.
- Professional development from business school faculty: The health system is sending the unit-based clinical leadership teams to attend the Penn Medicine Leadership Forum program designed by the University of Pennsylvania Health System's Penn Learning Academy and the Wharton School of Business. The program consists of four 8-hour days spread over several months. Attendees receive leadership training on topics such as idealized design, identifying stakeholders, project development and implementation, problem framing, and innovation. Participants work on an action learning team to put the skills learned into practice on an actual project aimed at improving transitions in care. The health system is also considering additional training for the leadership teams, including project management and/or quality certification.
- Messaging campaign for frontline staff: Unit-based teams are working together on a campaign to highlight the initiative to frontline staff and leaders within all hospital departments. Communication will occur via presentations at unit and system-wide meetings and through other vehicles.
- Grand rounds/quality exchanges: As noted, program leaders are developing cross-institutional grand rounds to facilitate learning and the sharing of best practices throughout the health system.
- Data warehouse: As noted, a data warehouse is being developed to facilitate the provision of real-time data to the teams.
- Performance improvement training: Information provided in June 2011 indicates that the teams are being trained on principles and tools related to the Toyota Lean, Six Sigma, and Plan-Do-Study-Act performance improvement models. The goal is to ensure a consistent approach to quality improvement across the organization. Program developers expect that 500 individuals will have undergone this training by mid-2012.
- New guidelines for physician participation: Information provided in June 2011 indicates that the hospital set new guidelines with standards and expectations for physician participation on the unit-based teams.
Resources Used and Skills Needed
- Staffing: The Hospital of the University of Pennsylvania hired two new quality/safety project managers to join the existing two specialists already employed by the organization. The other hospitals in the health system also employ quality specialists who support the unit-based clinical leadership teams. Information provided in June 2011 indicates that the hospital has hired an additional three project managers, for a total of five. Potential adopters might need to hire new staff to fill this quality specialist role, and ideally, staff who fill this role should have some direct patient care experience. Nurse managers absorb unit-based clinical leadership duties as part of their full-time managerial role. Physician leadership is expected to devote 20 hours monthly to the program. Information technology personnel also dedicate time, primarily related to data generation and analysis.
- Costs: The actual cost of the program for the first year was approximately $330,000. This is a conservative estimate based primarily on physician stipends allotted to each physician department to cover the time of physician leaders. Data on the annual operating costs of the initiative is unavailable.
Getting Started with This Innovation
- Begin at senior level: Once senior nurse and physician leaders begin to work together, the development of healthy relationships based on mutual respect will manifest on the unit level, as providers collaborate to implement improvement activities.
- Bundle units to create synergy: Some units can be combined into one team, typically those with similar patient populations and/or similar types of quality improvement priorities.
- Ensure budget support: Physician leaders' time on the program should be supported via the health system budget.
Sustaining This Innovation
- Share best practices across teams: Quality leaders should serve as liaisons between teams, facilitating the sharing of best practices and avoiding duplicative efforts.
- Share data on program impact: Sharing program outcomes helps build support among frontline staff and in other (nonadopting) units. At the University of Pennsylvania, the program's success has prompted leaders in other units to express interest in it.
- Do not get discouraged: Changing culture and improving nurse–physician collaboration takes time. Cultivate continued enthusiasm for the project by periodically highlighting patient safety as a common goal for administrators, physicians, and staff.
- Continue to enhance and refine the program: Information provided in June 2011 indicates that over time, program developers should continue to examine the processes used by and the potential roles of the unit-based clinical teams. Refine the teams by adding participants from additional disciplines, enhancing training, setting participation standards, and considering what new contributions they can make to health system clinical and operational quality.
- Be prepared for expansion: Information provided in June 2011 indicates that other units throughout the hospital will hear about the teams and proactively request to implement the team model.
Contact the Innovator
Note: Innovator contact information is no longer being updated and may not be current.
Patrick J. Brennan, MD Chief Medical Officer and Senior Vice President, Penn Medicine
Perelman Center for Advanced Medicine
3400 Civic Center Boulevard, A-5
Philadelphia, PA 19104-4283
Dr. Brennan has not indicated whether he has financial interests or business/professional affiliations relevant to the work described in this profile.
Robert Wood Johnson Foundation. University of Pennsylvania Pilot Project Targets Physician-Nurse Communication. August 18, 2008.
Arford PH. Nurse-physician communication: an organizational accountability. Nursing Economics. 2005 Mar-Apr;23(2):72-7, 55.
Interview with Kate FitzPatrick and Jeff Rohrbach, November 11, 2009.