Skip Navigation Archive: U.S. Department of Health and Human Services
Archive: Agency for Healthcare Research Quality
Archive print banner
Evaluation of AHRQ's Partnerships for Quality Program

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Appendix B: Summaries of PFQ Grantee Activities (continued)

PFQ Grant Summary: Closing the Gap: Partnering for Change

Lead Organization: American College of Physicians (ACP)
Partner Team: Northwestern University, Abington Memorial Hospital
Title: Closing the Gap: Partnering for Change
Topic Area: Process Continuing Medical Education to Improve Quality of Care
Principal Investigators: Vincenza Snow, MD
AHRQ Project Officer: Charlotte Mullican
Total Cumulative Award: $848,736
Funding Period: 9/2002-9/2005 (project funds not released until February 2003)
Project Status: Completed 9/29/2006

1. Project Description

Goals.  The aim of this project was to (1) develop and test a team-oriented, practice-based continuing medical education (CME) strategy that trains teams of doctors, nurses, and office administrators in how to improve quality of care and outcomes for patients with chronic diseases, and (2) design a business case that would help spread the adoption of team-oriented, practice-based CME by the ACP and other professional societies. The project team hoped to show that the new team-oriented, practice-based approach to learning would be a better way to promote physician adoption of clinical practice guidelines and improve quality of care for patients.  The team also intended to establish this type of CME as a viable alternative to traditional CME, which is physician centered and based on passive learning. For this trial, the prototype CME learning strategy focused on educating physician practices on type 2 diabetes care. 

Activities and Progress. In the first funding year (September 2002-September 2003), despite problems gaining IRB approval that delayed grant work by about six months, the project team established partnerships with key national stakeholders to create a project advisory board (see table for members). This group helped to design the education program and develop a training manual on learning collaboratives and a team-oriented toolkit for diabetes. Together, the training materials are called "Closing the Gap Diabetes Modules." 

The project also recruited four ACP practices in Pennsylvania and Illinois to participate in the pilot test of the practice-based learning model for diabetes.  The pilot test began in October of the second funding year (October 2003-September 2004).  Each practice that participated in the pilot project chose a team composed of one doctor, one nurse, and one administrator to attend three training sessions over a six- to nine-month period. One session was held on each of the following: performance improvement, the Plan-Do-Study-Act (PDSA) cycle, and the fundamentals of the Chronic Care Model.  During this time period, the teams returned to their practices to train other staff and implement the team-oriented diabetes toolkit, which included clinical, administrative, and patient tools intended to redesign practice workflow. In between the three training sessions, the primary program trainer, Dr. Kevin Weiss of Northwestern University, held two conference calls lasting two hours with the practices to keep them on track and guide them through operational changes. Information from the pilot practices' learning experiences, responses to barriers, and perceptions on how the team functioned differently informed revisions that the research team made to the trial intervention. 

Following the pilot study, during the third funding year (October 2004-September 2005), the research team began to implement the pseudo-randomized trial intervention. The team successfully identified and recruited 25 practices in Philadelphia (randomized into 13 intervention practices and 12 control practices) and 6 practices in Chicago (randomized into 3 intervention and 3 control practices) to participate in the study. Rather than conduct a true randomized trial, in which control practices would not receive training, the study design was changed to allow the control practices to receive the intervention as soon as the experimental practices completed the training program. This change was prompted by the insistence of one hospital system that had volunteered 25 internal medicine and family practices to the study and wanted all of them to benefit from it.  

The first training session occurred in October 2004, and the intervention proceeded as it had in the pilot, except that the three full-day training sessions were reduced to one full-day and two half-day training sessions, and the training materials were revised to include only the most relevant and useful ones. The research team designed an evaluation to measure three sets of outcomes: (1) patient outcomes and practice patterns, (2) patient satisfaction, and (3) practice teams' perceptions of the program. To collect patient outcomes and practice pattern data, each practice (both experimental and control) enrolled 15 patients with diabetes and extracted data on HbA1C levels, blood pressure levels, blood glucose, and lipid control from patient charts three times during the study.  The practices sent this data to the Data Coordinating Center at Northwestern, where it was cleaned, analyzed, and used to create reports for each practice on its patients' status at baseline, during the intervention, and afterward.

Data collection in the trial study was delayed by the slow pace of recruiting patients and extracting data from charts. By June 2006, however, the research team received all three rounds of data from the Philadelphia practices and about 80 percent from the Chicago practices. To collect patient satisfaction data, practices helped recruit patients with diabetes to participate in a telephone survey, which staff at Northwestern had planned to conduct three times during the study: before, during, and after intervention. However, because of problems enrolling patients in some practices, and errors in sending the correct consent forms to control groups in Philadelphia, the interviews were delayed. As of June 2006, the patient surveys were complete and researchers were analyzing the results. 

2. Partnership Structure/Function

There were three levels of partnership in this project.  The first involved ACP and Northwestern University, whose staff formed the core research team, including a project principal investigator from ACP and a co-investigator from Northwestern. This team spoke regularly and together designed the pilot test, the trial intervention, and the teaching materials. They also provided the training and support to practices, and collected and analyzed the data. The second partnership involved the ACP-Northwestern research team and the physician practices that participated in the pilot study and the trial intervention training programs. Practices had regular contact with Dr. Weiss at Northwestern, who provided them with ongoing technical support.

The third level of partnership involved the ACP-Northwestern research team and members of an advisory group, who provided input to the project's design and teaching tools (Institute for Heathcare Improvement [IHI], Institute of Chronic Illness Care [ICIC]), offered avenues to disseminate outputs from the project, and facilitated participation of practice-based health providers (American Medical Association [AMA], AHIP, American Nurse Association [ANA]). In the first year, the project had one in-person advisory board meeting at which members could cement relationships and reach agreement on a conceptual model of the team-oriented, practice-based diabetes prototype. The project also created working groups—one on the business case and another on implementation and barriers—composed of advisory board members and other key partners. The project held mini-strategic planning teleconference calls with the working groups to develop different modules of the training program.

Table 1. Major Partner Organizations and Roles in the Project



Role in Project

Lead Organization (grant recipient)

American College of Physicians
PI: Vincenza Snow, MD

Provided overall leadership and direction to program; guided the design of CME intervention and training materials; developed and implemented training programs and developed evaluation plan on project impact; assessed opportunities for expansion and sustainability of project outcomes

Key Collaborators

Northwestern University
Co-PI: Kevin Weiss, MD

Institute for Healthcare Improvement

Institute of Chronic Illness Care

American Medical Association and American Association of Health Plans

American Diabetes Association and American Nurse Association

Guided design of CME intervention and training materials; provided training and technical assistance to participating practices; collected and analyzed data for pilot test and randomized control trial

Participated in advisory board; assisted in developing training materials for practices (training manual on learning collaboratives—IHI; toolkit for diabetes—IHI and ICIC)

Participated in advisory board; assisted in identifying opportunities for dissemination of project outcomes and sustainability of project activities

Participated in advisory board; assisted in gaining participation from nurses by providing CE credit (ANA)

Target Organizations

Four practices from Pennsylvania and Illinois for pilot test (one was a Lehigh Valley practice in PA identified through another PFQ project)

31 practices (experimental and control); 25 in Philadelphia and 6 in Chicago for trial intervention

Participated in team-oriented, practice-based diabetes CME prototype; attended training sessions; participated in conference calls; implemented changes to practice workflow based on training; performed data extractions and sent data to the Data Coordinating Center; recruited patients for patient satisfaction survey

3. Project Evaluation and Outcomes/Results

The project successfully created a set of diabetes training modules for practice-based teams; pilot tested the module with 4 practices; recruited 35 practices (4 for the pilot and 31 for the trial) and patients from those practices to participate in the randomized control trial intervention; and gathered clinical trial data. While the research team did not have information at the time this summary was written (September 2006) on the impact of the program on patient clinical outcomes or patient satisfaction, it did complete the qualitative evaluation of the practice teams' program experience and level of collaboration.  The results showed that practices were willing to attend training in, learn from, and participate in the project's team-based learning model, in spite of the cost involved in sending three employees to training sessions.  The trial intervention had about an 85 percent participation rate from the experimental group, with 15 percent (about one to two practices) showing inconsistent participation.

The research team evaluated the practice teams' experience and the level of team collaboration with a pre- and postintervention survey of the practices. Despite the intensity of this program, participants rated it highly, while at the same time complaining of the high intensity. Over the three training sessions, 94 percent of participants rated the program as "very good" or "excellent." But "very good" to "excellent" ratings dropped from 96.7 percent of participants for sessions 1 and 2 to 88.2 percent for session 3, possibly reflecting fatigue. When asked what was the most "eye-opening experience" for them, participants rated "working as a team" as the highest followed by "interacting with the other teams," "learning improvement strategies," and "reviewing their charts."  The first two relate to the in-person meetings, but the team interactions were also part of the conference calls and could be accomplished via an on-line community. Program participants rated the binder contents as most useful to learning, and within these, the care models, patient tools, and chart tools were of greatest value. They also rated the conference calls highly as a learning experience. Participants rated measuring their practice, progress reports on the conference calls, and patient satisfaction data as having the greatest impact on their ability to improve practice, followed by the binder materials and the learning sessions.

The project was found to be helpful to nurses and office managers.  These practice staff indicated that the learning model helped to integrate them into the care process by opening up dialogue between physicians and staff. One physician practice noted that staff members felt a renewed sense of purpose because the project gave them tools for comanaging patients. Office managers often played a key role in the project at the practice level by keeping track of patients in the project. 

The project established "face validity" for the learning model with physicians.  Feedback and testimony from physicians were positive; some practices indicated that the program changed the way they practice by showing them the benefits of incorporating program tools, such as new forms and databases, into everyday workflow. For example, one practice introduced a scorecard that the nurse fills out with information on patient health status, diabetes care status, and instructions for self-care.  The practice gives a copy of the scorecard to the patient and keeps a copy from which to enter patient data into its computer registry to track performance over time. Other practices made changes in office procedures, such as having nurses help patients take off their shoes as a reminder to physicians to check their feet, or instituted new patient education initiatives. 

ACP and other organizations like the AMA and the American Board of Internal Medicine (ABIM) had positive reactions to the new CME model.  Partly due to the success of the ACP project, which was part of an AMA pilot to test practice-based CME, the AMA decided to award 20 category 1 CME credits to physicians participating in practice-based programs like ACP's Closing the Gap, and ACP is now accredited to provide practice-based CME.  In addition, ABIM now accepts participation in ACP's Closing the Gap as fulfilling part 4 of its requirements for Maintenance of Certification. The program was featured at an ABIM Quality Summit as a "premier project for the ACP in helping members achieve higher levels of quality care and become eligible for pay for performance projects" (ACP, Mid-Year Progress Report to AHRQ, June 2006). The ABIM considers Closing the Gap as the "gold standard" against which all other practice-based CME programs are measured.  ANA also approved CE credit for nurses involved in the program. Finally, many ACP state and local chapters, which were initially hesitant to participate in the study, are now anxious to do so.

4. Major Products

  • Closing the Gap Diabetes Modules, including a Manual on Learning Collaboratives for the practice teams, and a toolkit for diabetes care.
  • Summary report on the pilot test experiences and barriers.
  • Presentation of the project's experiences at the ACP's annual session in 2005.
  • News articles in ACP newsletters and electronic newsletters, distributed to 70,000 ACP members (
  • Patient data registries, scorecards, and other tools that practices created to track diabetic patients.

5. Potential for Sustainability/Expansion after PFQ Grant Ends 

ACP's Closing the Gap project led to larger projects that are further testing the team-oriented, practice-based learning model through follow-up pilots.  The project has received funding from two pharmaceutical companies to conduct two more rounds of Closing the Gap training programs, one in diabetes (funded by Novo Nordisk for $9 million) and one in cardiovascular disease, with 20 practices in each group. Several physicians who received training in the initial study have become faculty for the new Closing the Gap programs and will teach the training sessions for new practices. 

The research team is working to develop a sustainable business case and financing for the program. The two biggest costs to practices are those related first to measurement and workflow changes, and second to the time staff spends being trained.  For ACP to expand this program, it also needs to find external funding. One option involves ACP's charging fees for the program, supplemented by contributions from local and state partners of ACP chapters. ACP is also considering ways to build the program into its internal budget and create its own data coordinating center, but this would also require external funding.  Finally, researchers are considering the development of a web-based version of the program that would be less costly and time-consuming for physicians—a "Closing the Gap 101" to teach the PDSA cycle—as a way to disseminate it more broadly.  The more intensive training in this program would be the next step, a "Closing the Gap 102" that would concentrate on the practice improvement and measurement components.

Return to Appendix B Contents

PFQ Grant Summary: Improving Care for the Dying: Transforming Patients' Wishes into the Reality of High-Quality Palliative Care

Lead Organization: American Hospital Association (AHA), Health Research and Educational Trust
Partner Team: Three Pennsylvania-based hospitals/hospital systems and four hospitals/hospital systems based outside Pennsylvania (national)
Title: Improving Care for the Dying: Transforming Patients' Wishes into the Reality of High-Quality Palliative Care
Topic Area: Palliative Care
Principal Investigators: John Richard Combes, President and Chief Operating Officer, Center for Healthcare Governance, AHA
AHRQ Project Officer: Ronda Hughes
Total Cumulative Award: $1,282,703
Funding Period: 9/2002-9/2006
Project Status: Completed 9/29/2006

1. Project Description

Goals. This project sought to promote the establishment of hospital-based palliative care by creating centers of learning for other hospitals, and to accelerate the translation of research findings into improved quality and delivery of end-of-life care.  In phase I, the project planned to establish three palliative care learning centers at Pennsylvania-based hospitals to host site visits by other hospitals interested in planning and developing similar palliative care units.  In phase II, the project planned to expand the number of learning centers to hospital-based palliative care centers in other parts of the country, selected from among recipients of the AHA's Circle of Life Award.

Activities and Progress. The first six months were devoted to planning and developing the core curriculum of the site visits with the initial three learning centers in Pennsylvania: Geisinger Health System, Danville; Center for Palliative Care in Thomas Jefferson University's Department of Family Medicine and the Jefferson Health System, Philadelphia; and the University of Pittsburgh Medical Center.

Phase I began during the second half of year 1 and expanded into year 2.  The project aimed for each of the three facilities to accommodate five site visits the first year and eight site visits per year for years 2 through 4, for a total of 29 site visits at each.  During year 2, phase II began with the establishment of four national learning centers (Connecticut Hospice in Branford, CT; Detroit Receiving Hospital in Detroit, MI; Palo Alto VA in Palo Alto, CA; and St. John's Regional Health Center in Springfield, MO).  The four were chosen among AHA Circle of Life Award winners and finalists, and represented different types of settings for palliative care (i.e., VA hospital, safety net hospital, Catholic hospital, and hospices).

The lead organization (initially Hospital and Health System Association of Pennsylvania) recruited hospitals or hospital systems to participate in site visits and matched up visitors with the learning centers. The learning centers contacted the hospitals to schedule the site visit and to conduct a preliminary needs assessment, in which staff members were interviewed to assess their unique clinical and community situation, areas of interest, and palliative care goals.  During the visit, discussion was guided by the data gathered during these pre-site interviews.  The learning centers tailored the site visit curriculum and schedule to the visitors' identified needs.  After the site visits, the lead organization followed up with the visiting hospitals to assess the effectiveness of the site visit and provide ongoing support and technical assistance.

As of early October 2006, approximately 60-70 site visits had been conducted.  Site visits lasted a full day and were hosted by a team of professionals, including physicians, a palliative care project coordinator, nurse clinicians, hospital administrators, clergy, social service professionals, and volunteer coordinators.  Members of the host organization team provided tours of the facility, supplemented by formal and interactive presentations.  Each site visit included a presentation on how the research collected during the developmental stages in regard to challenges and successes was translated into improved palliative care services and procedures.  The host team encouraged visitors to share their research findings and solicit approaches to translating them into successful practices.  Discussions focused on how to ensure that systemic change, including policy change, occurred, and on how to create a supportive environment so that established palliative care services could be sustained.  Host organizations shared data used for benchmarking, internal and external marketing strategies, reimbursement and funding challenges, outcome measurements, evaluation process, and views of how systemic change holistically influenced the delivery of health care within their organization.

2. Partnership Structure/Function

During the initial planning phase, the three Pennsylvania-based hospitals/hospital systems spoke with the principal investigator (PI) by phone every other week and in person once per quarter to build the site visit curriculum.  The PI, project director, and seven learning centers (called "learning labs") did planning via conference calls held approximately every six weeks.  These conversations provided the team with the opportunity to evaluate the effectiveness of the program process, brainstorm on continued marketing and training strategies, and continue group discussion and work on collaborative projects such as survey development and refinement of curriculum and site visits. In addition, member listserves, the Hospital-Based Palliative Care Consortium Hospital-Based Palliative Care Consortium Web site, and conference calls facilitated communication between the lead organization, participating hospitals, and learning labs.

Table 1. Major Partner Organizations and Roles in the Project



Role in Project

Lead Organization (grant recipient)

Health Research and Educational Trust, AHA

To provide overall project leadership

To identify and recruit learning labs

To develop core curriculum for the site visits and companion toolkit

To recruit participating hospitals (through Web sites, electronic newsletters, learning lab institution publications, and various meetings and conferences)

To develop assessment tools to evaluate the usefulness of the learning labs for the visiting/participating hospitals

Key Collaborators

Phases I and II: Palliative care programs in 3 PA-based hospitals and hospital systems [Note: By the end of the grant, one of the PA-based learning labs had dropped out of the program.]

To assist in developing the core curriculum for the site visits and companion toolkit (Phase I hospitals/hospital systems only)

To conduct and assess pre-site-visit surveys filled out by the visiting hospitals/hospital systems

To coordinate and host site visits

Phase II: Palliative care programs based at 4 hospitals and hospital systems (national) To respond to follow-up questions/inquiries from visiting hospitals

Target Organizations

Hospitals and hospital systems throughout the U.S.

To complete pre-site-visit assessment

To visit learning labs and adapt evidence-based models of change to incorporate palliative care services into hospitals/hospital systems

3. Project Evaluation and Outcomes/Results

The program planned to evaluate its success according to the number of new hospital-based palliative care programs created in targeted hospitals,2 and the number of enhancements made to current programs as a result of the training program. About 60-70 site visits had been completed at the time this summary was written (October 2006).  Initially, the evaluation intended to measure outcomes such as reduced length of stay, patient and family satisfaction, and the financial effects of instituting hospital-based palliative care services.  However, the learning labs were concerned about measuring patient satisfaction.  Specifically, they felt that while those patients and families who participated in the palliative care program would report positive effects, patients and families who did not receive palliative care services might skew the results.  As a result, the three Pennsylvania pilot hospitals serving as learning labs provided only baseline clinical and financial data prior to the initiation of phase I.  During phase II, AHA-HRET staff surveyed state and national learning labs to evaluate the impact of the palliative care programs on these outcomes.  These data will be compared to the baseline data collected from the three Pennsylvania-based learning labs prior to phase I.

In addition, AHA-HRET staff conducted followup with visitors approximately six months to one year after the site visit to explore whether expectations were met, what was learned from the visit, what new services were developed as a result, how services were functioning, etc.  Project staff planed to analyze this information at the end of summer 2006 (as of October 2006, we were unsure if this was completed as scheduled).

The project has also produced less tangible but nonetheless important lessons.  For example, many hospitals have been reluctant to adopt the program because revenues are reduced if people spend less time in the hospital, even though use and cost of inappropriate services are also decreased.  One of the learning labs taught visitors how to capture allowable charges.  The project also found that each set of stakeholders—hospital CEOs, CFOs, physicians, and nursing staff—have different concerns that need to be addressed to gain their support for a palliative care program.

4. Major Products

  •     "Back to School: A Unique Education Program Provides Hands-On Experience with Palliative Care." Hospitals and Health Networks, November 2004.
  • Implementation of Hospital-Based End-of-Life and Palliative Care. Poster presented at AHRQ's 2004 TRIP Conference, July 12-14, 2004.
  • Recruiting-oriented presentations: American Academy of Medical Administrators, Boston, MA, November 2002; Partners for Quality, Rockville, MD, March 2003; Medical Advisory Board Lehigh Valley Hospice and Home Health, Allentown, PA, April 2003.

5. Potential for Sustainability/Expansion after PFQ Grant Ends

While there is no funding in place for sustaining this project, it is possible that some learning labs will continue to host scaled-down versions of the site visits, if approached by hospitals/hospital systems. It is also possible that something may arise from AHA policy leaders', concerns about the disproportionate amount spent on end-of-life care, AHA leadership have discussed support for palliative care as a way to reduce that spending but have not taken any steps towards this, other than the Circle of Life Awards.

2. A similar program, the Center to Advance Palliative Care (CAPC) at Mt. Sinai Hospital in New York, funded by the Robert Wood Johnson Foundation, used a similar approach to promote hospital-based palliative care programs. It targeted larger hospital systems and university-based hospitals, however, whereas this AHA-HRET program targeted smaller community hospitals, VA hospitals, and safety net hospitals.  Also, CAPC charged hospitals to participate in its learning programs, while AHA-HRET did not.

Return to Appendix B Contents
Return to Contents
Proceed to Next Section


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care