Appendix B: Summaries of PFQ Grantee Activities (continued)

PFQ Grant Summary: Effecting Change in Chronic Care: The Tipping Point

Lead Organization: American Medical Association (AMA)
Partner Team: Iowa Foundation for Medical Care (IFMC), Northwestern University, Cook County Bureau of Health Services, United Healthcare Group (UHC), Midwest Heart Specialists (MHS), Pittsburgh Regional Healthcare Initiative (PRHI) and others
Title: Effecting Change in Chronic Care: The Tipping Point
Topic Area: Improving care processes and outcomes for chronic conditions
Principal Investigators: Karen Kmetik, PhD
AHRQ Project Officer: Cynthia Palmer
Total Cumulative Award: $1,211,074
Funding Period: 9/2002–9/2006
Project Status: Completed 9/29/2006

1. Project Description

Goals. The goal of this project was to achieve a "tipping point" in quality improvement in caring for patients with chronic illness—specifically adult diabetes, coronary artery disease (CAD), and major depressive disorder (MDD)—by advancing the widespread use of physician performance measures in various settings.  The primary interventions/tools for the project are measures developed by the Physician Consortium for Performance Improvement, which is convened by the American Medical Association (AMA), and the National Diabetes Quality Improvement Alliance.

The project originally aimed to test two approaches to collecting data on physician performance. One would establish a regional data warehouse for pooling payer claims data (United, Blue Cross, and CMS) in the Pittsburgh area and allow physicians to retrieve the data to assess their own performance (the "Community Model"). The other involved the electronic transfer of data from physician offices and laboratories to a central data repository in the Midwest (the "Practice Model"). The project planned to examine the impact of the two models on improved care processes and outcomes, identify implementation issues and challenges, and determine what would be necessary both to roll out the models nationwide and sustain participation by key partners.  

Activities and Progress

Year 1. In the first year, the project standardized the performance measures and tools for diabetes, CAD, and MDD, and began to pilot test two different models that could be used to provide physicians with performance measurement data at the point of care. The Pittsburgh Regional Healthcare Initiative (PRHI), one of the project partners, began to test the Community Model, which would compile data from health plans, laboratories, and a QIO to "pre-populate" a community data registry for physician retrieval. The Iowa Foundation for Medical Care (IFMC), another project partner, began to test the Practice Model, in which physician practices would generate data and send the information to a QIO or health plan for quality oversight purposes. Both models used the agreed-upon standardized performance measures.  

PRHI secured the commitment of five primary care physician practices with a total of 111 physicians providing care to more than 250,000 patients to participate in the pilot test.  PRHI met with practicing physicians, health plans, and laboratories to identify data capabilities and then secured preliminary agreements from some payers for integrating data from multiple sources into a regional community database called the Pittsburgh Health Information Network (PHIN) overseen by the Pennsylvania QIO. Physicians would be able to access patient data stored in the registry in standardized reports.  

IFMC secured the participation of four cardiology practices, a family medicine practice, and an internal medicine clinic; 79 physicians from the cardiology practices agreed to collect data for the CAD measures and 22 family practitioners, and 2 PAs and 3 internists agreed to collect data for the MDD measures. An assessment of the practices' current data capabilities found that the practices were at various stages of implementing electronic health record systems (EHRS).  The four cardiology practices collected baseline data and initiated ongoing collection of patient data. 

Year 2. In the second year, IFMC's arm of the project progressed; the practices that used EHRS successfully integrated the CAD performance measures into their systems. The paper-based practice sites struggled to integrate data collection into routine care, highlighting the significant advantage afforded to EHRS users in entering and retrieving treatment data, and in managing the care of patients. Based on this experience, the project decided to focus exclusively on the collection and reporting of data electronically, either using a data registry or the EHRS. 

Year 3. Problems in implementing the PRHI community model that emerged in year 2 caused this component to be discontinued in the third project year.  Project leaders failed to secure the participation of the University of Pittsburgh Medical Center Health Plan and CMS to contribute to the data warehouse because of legal concerns about data privacy. Without these vital sources of clinical data for the registry, the PHIN was not likely to be widely used in the community. In addition, the project did not have enough financial resources to build the health information network, technical problems emerged in its design, and doubt arose about the usability of the system by physicians. 

While work on the Community Model ended, the project realized that expansion of the Practice Model would be needed to truly reach a "tipping point" in improving the care of patients with chronic illness. Thus, the project expanded its activities to (1) include more practice sites (e.g., community clinics) with different EHR systems, (2) demonstrate the validity of physician performance data collected, and (3) provide concrete examples of both the data extraction process from physician offices' EHRS and the exportation of the data to other private and public users.

Two new partners were brought on board to allow for this expansion in the project work.  The project partnered with Cook County Bureau of Health Services to conduct a disease registry pilot to show how quality measures could be integrated into a commercial electronic disease registry system (DocSite) that would allow for data collection, monitoring and improvement of patient care, and provision of population-based feedback reports to participating physicians and clinics.  Northwestern University came on as a partner to work on a data validity pilot to implement and validate heart failure (HF) measures for an existing commercial EHRS (EPIC).  Midwest Heart Specialists (MHS), a large cardiology practice that was already involved in the project, worked with IFMC and United Healthcare (UHC) to begin a data export pilot that involved extracting data from an EHRS and exporting it to IFMC and UHC, using the HL7 file format which has been endorsed by HHS and CMS as the federal messaging standard.

Year 4. The final project year focused on publication of results from the performance measures testing, validation work, and other implementation efforts, as well as meetings to discuss the significance of the work and how it could be sustained through, for example, the AMA's Cardio-Health Information Technology (HIT) project.

2. Partnership Structure/Function

AMA served as the leader or convener for this partnership, which involved many different organizations over the course of the four-year project.  Partners included payers (United Healthcare, CMS, and BCBSA), physician groups, a QIO, a community health care coalition, an employer health coalition, and a county-based system of ambulatory care clinics.  AMA organized the partners' resources into one or more of the project components and contracted with some partners to support the work.  AMA also convened all-partner meetings via monthly phone calls as well as annual in-person meetings to share progress reports and lessons learned. As the project progressed, all-partner phone calls continued to occur at least quarterly but have begun to taper off as project activities began to wind down and partners became involved in spin-off projects. 

Of the initial project partners, PRHI and MBGH ended their involvement in the project either because their part of the work came to an end or the organization's priorities changed. United Healthcare, Northwestern University, and Cook County's Bureau of Healthcare Services came on as partners in later years of the project as work expanded.

Table 1. Major Partner Organizations and Roles in the Project

	Organization	Role in Project
Lead Organization (grant recipient)	American Medical Association (AMA)	Lead and coordinate the project, provide the evidence-based performance tools and interventions
Key Collaborators	Pittsburgh Regional Healthcare Initiative (PRHI)—ended participation when the Pittsburgh Health Information Network failed to become operational Iowa Foundation for Medical Care (IFMC) (QIO for Iowa and other states) United Healthcare/Ingenix CMS and Blue Cross and Blue Shield Association	Regional partner that served as the lead for testing the Community Model; identified and recruited physician practices to participate in pilot testing Regional partner served as the lead for testing the Practice Model; tested information tools for CAD and MDD; identified and recruited physicians practices to participate in pilot test; involved in data export pilot with MHS and UHC Involved in data export pilot with IFMC and UHC "Connectors" to other organizations to promote dissemination grantee efforts
Target Organizations	3 ambulatory care practices or networks in the Chicago region: Midwest Heart Specialists (MHS); Northwestern University General Internal Medicine/Medical Faculty Foundation; Ambulatory and Community Health Network/Cook County Bureau of Health Services	Participate in the pilot tests of information technology and tools to assess adherence to performance measures for chronic diseases

3. Project Evaluation and Outcomes/Results

The project learned through the Practice Model that getting physicians to use performance measures to improve care worked best in practices that had an existing EHRS.  According to the RAND evaluation report after the third project year, AMA came to recognize that achieving a "tipping point" in advancing widespread use of physician performance measures requires (1) involving more practice sites in collecting data through different types of electronic health record systems, (2) demonstrating the validity of physician performance data that are collected, and (3) showing how the data extracted from physician office EHRS can be easily exported to a wide array of public and private users.  

The interim RAND evaluation report, however, stated that the project's experience has not addressed some challenges faced by physician practices that want to take advantage of current technologies to measure their performance against AMA quality standards: (1) how to incorporate the Consortium's measures in physician office-based EHRS so that data on the measures can be generated by the system, and (2) exporting the data to a health plan or other party in a useable fashion.  The evaluation indicated that the experiences of the three pilots—disease registry, data validation pilot, and data export pilot—are inconclusive on both these issues. However, physician offices and clinics using the Consortium's measures in EHRS and disease registries report that they have seen, at least to some degree, process improvements and positive patient outcomes.  While these results cannot be definitively attributed to use of the Consortium's measures, the RAND evaluation concluded that it is reasonable to believe the measures had at least some positive marginal impact. 

Individual results from the three pilot studies include:

• Data export pilot—After experiencing difficulty with the data format, MHS successfully transferred a data file with "dummy" clinical performance data. One of the organizations receiving the data viewed the pilot project as successful since it demonstrated the ability to export clinical performance data from an EHRS to a QIO.  However, the other organization that received the data did not view the pilot project as positively due to the problems it encountered with the format of the data that was transferred, which made it less useful to them. 
• MHS successfully integrated Consortium measures into home-grown EHRS and has begun to provide tracking reports from data collected on Consortium measures to practice physicians. Validity testing for the Consortium measures was ongoing and a manuscript of results was in development as of June 2006. 
• Data validation pilot—The Northwestern team has been able to integrate Consortium measures into its commercial EHRS and generate performance data using HF measures and CAD measures.  Northwestern has been focused on educating their physicians on how to document and enter patient information into the system and are working on process and workflow redesign. Eventually, they hope to provide physicians with performance reports. Northwestern's validation work has helped the AMA refine its sets of HF and CAD measures. Two papers on the results of the validation pilot have been written and submitted for publication.  
• Disease registry pilot—Cook County has integrated the Consortium's asthma and the Alliance's diabetes measures in a commercial electronic disease registry.  Participating ambulatory clinics have begun to use the measures to do population-based care management. While the measures have been fully integrated in the disease registry, inputting necessary data into the system remains a work in progress.  The RAND evaluation indicated that the measures have positively impacted physicians in the nine participating primary care clinics. Many of the physicians report that the registry has helped them provide higher level of care, as evident in improving performance measures, decreasing number of patients in the high-risk group and increasing number of patients in the low-risk group.  Cook County is working to link its disease registry in its ambulatory setting to its commercial EHRS in its inpatient setting.

Another important result of the project was a June 2006 meeting convened by AMA with 25 electronic medical record vendors, CMS, and a Northwestern co-investigator to discuss improvements that could be made to electronic health record systems and products, which would make it easier for physician practice use.

4. Major Products

• O'Toole MF, Kmetik KS, Bossley H, et. al.  Electronic health record systems: the vehicle for implementing performance measures. Am Heart Hosp J. 2005; 3:88-93.
• Two papers written by Northwestern that have been submitted for publication. 
• A paper being written by MHS. 
• A paper being written by Cook County RAND's third-year evaluation of the project.

5. Potential for Sustainability/Expansion after PFQ Grant Ends

The success of MHS in implementing the Consortium's CAD measures in an EHRS launched a follow-on project called "Cardio-HIT—Physicians Advancing HIT to Improve Care", which was also funded by AHRQ and led by the AMA and MHS. The three-year project plans to spread the MHS model to six other physician practice sites in four different regions, using different EHRS systems. The project hopes to establish a data warehouse to enable feedback reports and benchmarking to support physician-directed quality improvement.  The seven practices will also work to integrate other Consortium measures into their systems. AMA also recently received a two-year grant from the Physicians Foundation for Health Systems Excellence, to continue working with MHS and Northwestern and add four more sites, each with different electronic record systems. Thus, the partnerships established between the AMA, Midwest Heart Specialists, and Northwestern will continue with these two projects. 

Return to Appendix B Contents

PFQ Grant Summary: Long Term Care Quality Improvement Partnership

Lead Organization: American Medical Directors Association Foundation (AMDA-F)
Partner Team: Quality Partners of Rhode Island; 20 national organizations represented in the National LTC Quality Coalition, and state or local chapters
Title: Long Term Care Quality Improvement Partnership
Topic Area: Improve implementation of AMDA Clinical Practice Guidelines for pain management and pressure ulcer reduction in long-term care (LTC) nursing facilities
Principal Investigators: David Polakoff, MD, MSc, CMD, Senior Vice President and Chief Medical Officer, Genesis HealthCare Corporation. Co-PI is David Gifford, MD, MPH, formerly with Quality Partners of Rhode Island, the QIO support center for CMS' nursing home quality improvement initiative, and currently Director, Rhode Island Department of Health
AHRQ Project Officer: Judy Sangl, ScD
Total Cumulative Award: $1,299,164
Funding Period: 9/2002–9/2006
Project Status: Completed 9/29/2006

1. Project Description

Goals. This project sought to determine the effectiveness of an approach for training nursing home staff to implement clinical practice guidelines developed by the American Medical Directors Association (AMDA), and to evaluate nursing homes' experiences and lessons learned in using implementation toolkits. The specific goals of the project were to  (1) develop a Long-Term Care Quality Improvement (LTC-QI) partnership that will enhance the quality of care and quality of life for nursing facility residents; (2) create national and local partnerships with LTC professional organizations, Quality Improvement Organizations (QIOs), long-term care facilities, and a national research network of more than 200 nursing facility medical directors to disseminate toolkits that translate AMDA clinical practice guidelines (CPGs) into practice; (3) identify and train interdisciplinary educators and mentors in six states to provide onsite CPG and CPG toolkit implementation training for 5 to 10 nursing facilities in each state (50 total); (4) collect and/or analyze data on both process and clinical indicators in the participating facilities to determine the effectiveness of the CPG implementation model and identify how it can be replicated independently in nursing homes; and (5) disseminate the model and refined toolkits in both online and print versions.

Activities and Progress. During the first year, the project created the National Quality Coalition, consisting of 15 partners, including representatives of nursing home associations (AHCA and AAHSA), the national QIO association (AHQA), AMDA members, and other key stakeholders. The Coalition advised the project on criteria for nursing homes participating in the project, strategies to recruit nursing facilities, which states to target, and other key design and implementation issues.  Six states were selected for the project: California, Florida, Indiana, Ohio, Pennsylvania, and Texas.  

The project leadership team (the PI and Co-PI, AMDA Foundation staff, and Quality Partners of Rhode Island) selected two CPGs—pain management and pressure ulcer reduction—as the focuses for CPG implementation. These clinical topics had been targeted for nursing home improvement nationally by CMS and were publicly reported on CMS' Nursing Home Compare Web site. Quality Partners helped to develop a plan for project implementation, specified indicators of CPG implementation, selected data elements for program evaluation, and created a "readiness matrix" to select participating nursing facilities. 

In the second year, facility recruitment began, and the selected long-term care facilities designated project teams consisting of the nursing home administrator, medical director, director of nursing, a data liaison, and others.  These teams participated in short (one day or less) training programs, run by state nurse consultants who were themselves trained by the AMDA Foundation Project Coordinator and Quality Partners staff.  Training consisted of review of the two guidelines, and guidance on how to initiate and manage organizational changes to promote adherence. AMDA developed CPG implementation toolkits that included sample letters/memoranda to staff.  The implementation training program was piloted during the 2004 AMDA symposium, and the implementation program and CPG toolkits were piloted with six facilities in Pennsylvania. In the pilot state of Pennsylvania, CPG implementation training was provided jointly to all participating teams, but in other states, nurse coordinators provided (to the extent possible) facility-specific training sessions for staff teams.  

The project team encountered unexpected problems and delays in recruiting facilities, which led to the loosening of some participation criteria, extension of recruitment areas to entire states rather than metropolitan regions, and allowing "rolling" enrollment. The project developed a web-based data reporting system and began collecting baseline data from participating facilities. Data on the CPG implementation process were to be collected at 11-and 18-weeks post-training, whereas data on clinical measures were to be collected at baseline, and at 9- and 15-months post-training. 

Program staff and partners in the National Quality Coalition made efforts to marshal support from state and local chapters of the national organizations to assist change in participating facilities, but generally were not successful due to limited capacity on the part of state and local chapters. By the beginning of the fourth year (October 2005), 54 facilities had been recruited, but some dropped out before receiving training or submitting baseline data, and others withdrew from the study due to changes in management or failure to submit follow-up data. In April 2006, 40 facilities were formally enrolled in the project and are expected to submit data for the evaluation. 

2. Partnership Structure/Function

The project leadership team included AMDA, AMDA Foundation and its Research Network, and Quality Partners.  The team held frequent conference calls and meetings.  The National Quality Coalition had annual meetings and, in the first year or two, quarterly conference calls, during which they provided input to the Leadership Team on project design issues. On a more informal basis, they communicated with state chapters and affiliates about the project, identified individuals in the selected states to serve as trainers; and provided forums at their national or state meetings to educate members about the project and recruit facilities for participation.  The national partners also disseminated information about project activities through publications, Web sites, and listservs. The original plan called for the identification of existing state and local coalitions to assist with recruitment of facilities and support dissemination of the toolkits and CPGs once the study was complete. Existing coalitions (or 'ready' coalitions) were to be identified in each of the six states, and were to play an active role in each phase of the study. Instead, only a few isolated local chapters of the national organizations in some of the states offered assistance to the participating facilities and teams. 

Table 1. Major Partner Organizations and Roles in the Project

	Organization	Role in Project
Lead Organization (grant recipient)	AMDA Foundation (Janet Pailet, Project Director)	Overall grant management; coordinate implementation of activities at the local level, including CPG implementation; create communication and dissemination plan
Key Collaborators	AMDA AMDA Foundation Research Network American Health Quality Association Quality Partners of Rhode Island National LTC Coalition (15 partners)	Provide clinical and executive leadership; work with CPG Steering Committee to create toolkits for pain and pressure ulcers; foster local partnerships Support for evaluation component (implementation and data collection at facilities) Liaison to QIOs—provide info about project and facilitates participation; holds forums for training and disseminating info Subcontract for Technical Assistance; oversee evaluation and analysis of implementation in participating facilities Advise the project on criteria for nursing homes participating in the project, strategies to recruit nursing facilities, which states to target, and other key design and implementation issues
Target Organizations	40-50 nursing homes in 6 states (CA, FL, IN, TX, OH, and PA)	Receive CPG implementation training and submit data to evaluate changes in processes of care and outcomes, as well as resource utilization

3. Project Evaluation and Outcomes/Results

The project collects process of care data through a web-based system and examines clinical outcomes. A separate, non-web based data collection effort gathers information about the CPG implementation process, including the amount of staff time spent on different tasks, the number of staff on the implementation team, compliance with each phase or component of the implementation process, and usefulness of the toolkit elements. No preliminary results were available when this summary was written (October 2006).    

4. Major Products

A manuscript, "Strategies for overcoming barriers to recruitment and enrollment of nursing homes in a national clinical practice guideline (CPG) implementation study" is in final preparation, and plans include manuscript development after data analysis is completed.  Articles about the project and the pilot states appeared in state LTC association newsletters, trade journals and newsletters, and a few local newspapers.  Project staff also wrote and issued a monthly e-mail newsletter, distributed to about 35 individuals and organizations, including those on the National LTC Quality Coalition.

5. Potential for Sustainability/Expansion after PFQ Grant Ends

The CPG implementation process is designed to be sustainable, in that the intervention involves only a modest amount of initial training and consultation by the state nurse coordinators.  For facilities that wish to implement CPGs, AMDA sells an implementation manual, which is available to any nursing facility at a modest price. But the motivation for using the CPG implementation manual and toolkits depends on evidence showing that their use contributes to tangible improvements in quality of care measures.  Those who received training to be CPG implementation trainers also may be resources for the state QIO or other nursing homes that wish to utilize their expertise.  Those QIOs that were involved in the project in the six states are more likely to promote this approach as part of their overall nursing home quality improvement activities. 

The National Quality Coalition established by the project involves organizations whose mission includes promoting quality of care improvements in long-term care facilities. Although the coalition itself may or may not last beyond the end of the project, communication and coordination among the members are likely to continue regarding related activities.  At the end of the AHRQ grant period, the project was testing the feasibility of transitioning the NQC to a Research Advisory Board for the AMDA-Foundation Research Network.

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