Appendix B: Summaries of PFQ Grantee Activities (continued)

PFQ Grant Summary: A National Center for Value Purchasing Models

Lead Organization: HealthFront
Partner Team: Park Nicollet Institute; National Institute of Health Policy; Colorado Business Group on Health; Buyers Health Care Action Group
Title: A National Center for Value Purchasing Models
Topic Area: Performance Incentives
Principal Investigators: Michael Callahan, former Executive Director at HealthFront
AHRQ Project Officer: Michael Hagan
Total Cumulative Award: $1,281,576
Funding Period: 9/2002–9/2006
Project Status: Completed 9/29/2006

1. Project Description

Goals.  The grant had two initial aims: (1) to develop a nationally recognized provider performance measurement, analysis, and award program, supported by purchasers; and (2) to develop the analytical capacity needed to support purchaser decisions on health care value purchasing. The grantee, HealthFront is a non-profit spin-off of the Minnesota-based Buyers Healthcare Action Group, with a board consisting of employer purchasers, health care consumers, and providers.  When another organization that was supposed to work on the first aim withdrew from the project, the grantee focused solely on the second aim.  Specifically, its goal was to evaluate methods for accelerating the adoption of "best practice" payment incentive systems by all major purchasers in selected communities by: (a) informing purchasers about the current use of incentives in pay-for-performance (P4P), public reporting, and tiered network strategies; (b) educating them about how to use incentive strategies; and (c) helping health plans align their respective incentives for P4P and public reporting.

Activities and Progress. Early in the first year after the project decided to focus on demonstrating how value purchasing could be supported and improved, the research team, comprised of researchers and staff from HealthFront, the National Institute of Health Policy, and Park Nicollet Institute, chose the Minnesota market for its initial test. The project partnered with the National Institute of Health Policy, led by former Senator David Durenberger and based at the University of St. Thomas (MN), and the Buyers Health Care Action Group (BHCAG), a group of major employers in the Minneapolis-St. Paul region that gave the project access to local purchasers and health plans.  In the first year, the project conducted interviews with about 65 health plans and provider organization representatives regarding their current use of incentives and measures for P4P and public reporting.  Results from these interviews indicated that there were vast differences among plans in their P4P activities and in the measures they used. The project team reported this information to purchasers to prompt discussions between them and the health plans about creating greater consistency in P4P and public reporting.

Due to other priorities, BHCAG did not follow up, but they have remained active with the Smart Buy Purchasing Alliance (a group of state and private health care purchasers). The core membership of the Alliance consists of a group of purchasers originally brought together by the grantee to discuss alignment of incentives.  Both BHCAG and HealthFront representatives serve on the Smart Buy Alliance. The Alliance recently made its first Bridges to Excellence physician bonus awards. Also, because of the state's involvement with the Alliance, the Minnesota Department of Human Services is pursuing incentive payment reforms for Medicaid hospital services based on advice from the project team. 

In the second year, the project work expanded into the Colorado market. The project partnered with the Colorado Business Group on Health (CBGH), which served as the conduit to employer purchasers in that community, and again conducted an assessment on the current status of P4P and public reporting in the market through interviews with local health plans and providers. The grantee presented the results of the assessment to purchasers, health plans, and other stakeholders.  Although interesting to stakeholders, the findings did not spark extensive dialogue between purchasers and health plans, nor did it lead to quantifiable action to align performance incentives. However, the CBGH credits the project with setting the groundwork for the community's entrance into Bridges to Excellence, a non-profit organization that recognizes and rewards health care providers for delivering quality health care.

In the third year, after the community assessments in Minnesota and Colorado were completed, the grantee brought together an expert panel via the Internet to discuss the role of incentives in improving preventive and chronic illness care, and the clinical capacity to manage care for better outcomes (e.g., registries, IT).  Providers and purchasers from the two communities also participated in the discussion. In October 2004, the project conducted a one-day in-person, retreat at the request of several of the panel members.

The panel, which included such experts in the area of quality effects of incentives as Robert Berenson, Lawrence Casalino, and Judith Hibbard, participated in the discussions, as well as small group exercises that identified the best ways for purchasers to provide incentives to providers.  These results were presented to purchasers in Minnesota and Colorado.  

One of the findings from the expert panel discussions was that communication was poor between medical practice leadership and rank and file physicians regarding P4P practices and public reporting. Since physician response to incentives determines the effectiveness of P4P, the grantee and partners, at the request of the purchasers, decided to obtain more information about what physicians know or think about P4P, public reporting, the use of incentives, and how they would respond to incentives.  Thus, in the third year, the project developed a survey for medical group managers in Minnesota to assess their perceptions of P4P, public reporting, and quality incentives in general.  Analysis of the survey results focused on responses from the managers of 78 unique medical groups representing 6,964 physicians in primary care practice in Minnesota.

In the fourth year, results from the survey were presented to purchasers and plans in the state, which generated substantial interest.  One of the findings was that a large number of physicians were uncertain about P4P and public reporting, either because they had a wait-and-see attitude or because they did not know much about it.  This suggested the need to educate physicians. The research team wishes to contact the physicians in Minnesota again to see if there have been any changes in plan activities (e.g., education activities for physicians) as a result of the findings.   

At the time this summary was prepared, the research team was fielding the physician survey in Colorado. Because practices in Colorado are smaller than those in Minnesota, the survey was revised to focus on the individual physician level rather than the group level.  Once the survey and the data analysis are complete, the project will present findings to the Colorado Medical Society at its annual meeting. The survey was supported by the local leaders of Colorado Medical Society, the Colorado Academy of Family Medicine, the American Academy of Pediatrics, and the American College of Physicians. 

2. Partnership Structure/Function

Project staff at HealthFront formed a core research team with two other groups: (1) health services researchers from Park Nicollet Institute, which is associated with a large multi-specialty medical group; and (2) the National Institute for Health Policy (NIHP), which is affiliated with the University of Minnesota and the University of St. Thomas.  (The former Executive Director of NIHP is now at the University of St. Thomas Center for Business Excellence but remains a key research partner in the project.)  Researchers from the three organizations held weekly meetings to develop and implement the surveys, conduct community assessments, analyze survey results, and plan for the dissemination of findings to community stakeholders.  

The core partners also formed partnerships with CBGH and BHCAG to gain access to purchasers in the community. The two purchaser coalitions hosted in-person meetings for the project team to present findings from the assessment of community activities in P4P, public reporting, and tiered network strategies. The team formed a close relationship with CBGH in Colorado, and the director of the purchaser coalition was actively involved in interviewing community stakeholders and analyzing the data. Relations with BHCAG in Minnesota were not as close because the organization was more focused on national issues.     

Table 1. Major Partner Organizations and Roles in the Project

	Organization	Role in Project
Lead Organization (grant recipient)	HealthFront	Responsible for project administration, coordination, research support, and employer liaison Assessed current state of P4P, public reporting, and tiered networks in Minnesota and Colorado through interviews with health plans and purchasers Reported on information from physician survey in Minnesota to purchasers and health plans to solicit stakeholder reactions and feedback
Key Collaborators	Park Nicollet Institute (PNI), Director, Health Systems Studies David Knutson National Institute of Health Policy (NIHP), Exec. Dir. Daniel McLaughlin Colorado Business Group on Health (CBGH) Buyers Health Care Action Group (BHCAG)	Health care services research center conducted research and survey design, financial analysis, and economic research, and was liaison with CMS and national research community Developed physician surveys, fielded surveys, and analyzed findings Participated in meetings to present findings from survey to stakeholders in MN University-based health policy research center (affiliated with University of St. Thomas, MN) provided liaison with CMS, health plans, Medicaid programs, policy, and educational institutions Helped gain access to health plans and other stakeholders for interviews to assess the status of P4P, public reporting, and tiering in Minnesota Hosted expert panel meetings to discuss findings and future steps for research; helped to analyze findings Helped access stakeholders in the market, including health plans, purchasers, and physicians Participated in interviews with stakeholders and helped to analyze findings Hosted the meetings to present information from assessment to CO community Hosted the meetings to present information from assessment to MN purchaser community
Target Organizations	Purchasers, health plans, physicians in the Minnesota health care market (in 2 areas: Minneapolis/St. Paul and rural western Minnesota) Purchasers, health plans, and physicians in the Colorado health care market (Denver)	Purchasers, plans, and physicians were interviewed by project staff to assess the community incentive environment in these markets Received information from the project's assessment of incentive environments Physician groups were surveyed for their perceptions on the use of incentives

3. Project Evaluation Outcomes/Results

Information from the community assessments was presented to purchasers and plans in each market. However, the information did not prompt discussions about value-based purchasing between purchasers and plans.  Although health plans in both communities are now working to achieve more consistency in measures used for P4P, public reporting, and tiered strategies, the work is not the direct result of the project findings.  In both Colorado and Minnesota, purchaser groups decided to work through the Bridges to Excellence program, rather than directly with health plans. In Colorado, however, project partners believe that grant activities contributed to the community dialogue that led to its decision to participate in the Bridges to Excellence program.

Researchers believe that information from the physician surveys on how they respond to payment incentives has the potential to affect purchaser behavior regarding value-based purchasing.  Particularly in Colorado, where the implementation of incentive programs was less advanced, the fact that employers are now engaged in an active dialogue with the medical community regarding value-based purchasing is directly attributable to the project. This dialogue, in turn, creates employer demand for such programs to be introduced by insurers and the discussion facilitates and informs implementation of these programs by educating the providers.  The plan is to follow up to determine to what extent purchaser or health plan activities can be attributed to survey information. The Colorado physician survey was completed by August 2006 and the results were presented in September 2006 at a meeting of the Colorado Business Group on Health, and at the Annual Meeting of the Colorado Medical Society. Both the employer members of the CBGH and, the leadership of the Colorado Medical Society in particular found the results of the survey enlightening. Researchers are drafting papers for submission to a peer-reviewed journal to include discussion of (1) the purchaser response to information on value purchasing, (2) results of the medical group manager and physician surveys, and (3) an exploration of the relationships between market penetration, alignment of incentive programs, and provider perceptions of them.

4. Major Products

• Medical group manager survey tool.
• Physician survey tool.
• Research findings regarding the responses of large and small medical groups to quality incentives, and recommendations from the provider community about desirable and actionable design features of quality incentives.
• Summary of an expert panel discussion that identified the best ways for purchasers to provide incentives to providers, and potential unintended consequences that plans and purchasers policymakers need to guard against. 
• Presentation of physician survey results to Colorado Medical Society, September 16, 2006.

5. Potential for Sustainability/Expansion after PFQ Grant Ends

Purchasers in Minnesota, including the Buyers Health Care Action Group, have expressed interest in having the researchers conduct a second round of the physician survey. The National Business Coalition on Health, a national non-profit membership organization of employer-based health coalitions, has expressed interest in working with the project's researchers to disseminate information to support its member coalitions in trying to improve quality through P4P, public reporting, and tiered network strategies.  The Colorado Medical Society has asked the team to write articles for its member publications and is interested in working with the researchers and the CBGH to continue the dialogue with physicians. The project team plans to conduct mini-case studies of local markets, how purchasers are using incentives, and how providers respond to them. The team is developing an online course on pay-for­performance directed toward an audience of physicians and medical group managers to be offered by the University of St. Thomas.  This online course builds on the team's experience with the online expert discussion panel sponsored by the University in 2004.

Return to Appendix B Contents

PFQ Grant Summary: Real-time Optimal Care Plans for Nursing Home Quality Improvement

Lead Organization: International Severity Information Systems, Inc. (ISIS)
Partner Team: IFAS/AAHSA, AHQA, Catholic Health Partners, Good Samaritan Society, National Church Residences, Christian Home and Rehabilitation, Sugar Creek Rest, Marywood Nursing Center, Ozanam Hall, Memorial Hermann Spring Shadow Pines
Title: Real-time Optimal Care Plans for Nursing Home QI
Topic Area: Improve prevention of pressure ulcers in nursing homes
Principal Investigators: Susan Horn, VP for Research at ISIS and Senior Scientist, Institute for Clinical Outcomes Research (ICOR is a division of ISIS). Co-Investigator is Robyn Stone, Exec. Director of the Institute for the Future of Aging Services/AAHSA in Washington DC.
AHRQ Project Officer: William Spector (originally Thomas Shaffer)
Total Cumulative Award: $1,297,577
Funding Period: 10/2002–10/2006
Project Status: Received a no-cost extension to March 2007

1. Project Description

Goals. This project incorporated research findings from the National Pressure Ulcer Long-term Care Study (NPULS) (1996) into routine, evidence-based best practice in long-term care (LTC) facilities. The project standardized front-line documentation and used this information to produce weekly reports to support clinical decision-making and care planning.  Through a staged approach, the project facilitated clinical process and workflow redesign, introduced technology tools that assisted providers in identifying high-risk residents, and empowered front-line staff to take appropriate and timely prevention or treatment actions.  Ultimately, the project aimed to redesign clinical workflow—instead of concentrating on improving existing processes only—to reduce the incidence of pressure ulcers among LTC residents in nursing homes.  

Activities and Progress. The project leadership team was led by ISIS; the co-PI at IFAS/AAHSA was involved in overall project assessment and promotion of project activities.  The American Health Quality Association (AHQA) provided assistance with dissemination of information regarding project activities, including presentations at AHQA national meetings and contact with the editor of the Provider publication. 

In the first year, the project selected a pilot site, Memorial Hermann Spring Shadow Pines in Houston, TX, which formerly had worked with ISIS on the NPULS project. Project staff designed scannable, comprehensive documentation forms for Certified Nursing Assistants (CNAs) and tested them at one nursing unit in the pilot site. AAHSA's Institute for the Future of Aging Services took the lead in recruiting and screening additional nursing homes for participation in the project, and ISIS used various networks to recruit study participants, including some affiliated with a PFQ grant recipient in Ohio. By April 2003, five additional nursing homes in four states had been selected and had agreed to participate.  By May 2004 (the second year of the project), 20 units in 12 nursing homes from 10 states had been selected to participate. The project began instituting systems to streamline documentation for CNAs and nurses. For CNAs, multiple logbooks, clipboards, and notebooks were consolidated into a single documentation instrument that included meal and fluid intake, weight, bowel and bladder incontinence, and behavior observations.  Nurses consolidated information into a CareGiver Guide that included pressure ulcer risk factors, medications, nutritional supplements, and fluid intake. ISIS assisted with facility-requested customization of the standardized forms. Clinicians used optical character recognition (OCR) forms, which allowed facility staff to use the familiar method of documenting on paper, and faxed them to ISIS where software exported the data to a database. ISIS generated weekly facility-specific reports and provided help with report interpretation to follow clinical best-practice guidelines at each facility. It also collected baseline data for evaluation, and began developing plans to sustain the process at the facility and unit levels.    

In the third year, the project held its second and third project meetings (November 2004 and April 2005); most participating facilities sent one or more representatives to share progress, challenges, and outcomes.  Many facilities expanded the use of CNA documentation forms to additional units, and some used the forms facility-wide.  Completeness rates varied; some facilities were very high (rates of more than 95%) and others were lower (50 to 60 percent). Facilities shared experiences with comprehensive documentation and gradually decided to use the same documentation forms, so that standardization was achieved.  The standardized CNA form replaced other forms and became part of the resident's medical record at each facility.  Most facilities began to incorporate data from the six ISIS-generated reports on resident status into daily or weekly resident care planning, which allowed staff to identify triggers for specific protocol steps to reduce the risk of pressure ulcers.

During the last year of the project, the focus shifted to sustaining project activities in participating facilities. ISIS helped facilities to explore ways of managing/sustaining process improvements without ISIS support, as for example through electronic medical records or digital pen technology. (See below, under Potential for Sustainability/Expansion.)

2. Partnership Structure/Function

The project formed an Advisory Committee to provide input and guidance on standardized documentation, implementation approaches, and analysis of results. Members included representatives from AMDA (medical directors of LTC facilities), academic researchers, a foundation representative, and the executive of a health care IT company.  In addition, the project organized a Working Group, comprised of representatives of participating nursing home sites, and including some combination of the facility's medical director, Director of Nursing, administrator, and MDS coordinator. According to a grantee report: "Another layer of partnerships exists within each facility. Each facility convened a QI team that is multi-disciplinary and includes all members of the care team, i.e., administrators, nurses, nursing assistants, social workers, MDS coordinators, dieticians, etc. This representation of all, especially front-line workers, is an atypical approach to QI efforts." The first project meeting included Advisory Committee members and facility representatives.

Table 1. Major Partner Organizations and Roles in the Project

	Organization	Role in Project
Lead Organization (grant recipient)	International Severity Information Systems, Inc. (ISIS)	Project management; convening Advisory Board and Working Groups of participating facilities Support to each participating facility to develop and process forms for each resident, generate reports, work with staff at all levels on implementation of facility-specific work plans Lead effort to sustain project activities
Key Collaborators	Institute for the Future of Aging Services/ AAHSA American Health Quality Association	Project guidance and support for establishing partnerships with project sites; recruit and screen project sites Provided assistance with dissemination and outreach for project activities, including presentations at AHQA national meetings and contact with editor of the Provider publication; also was a conduit to key leaders of nursing home trade associations
Target Organizations	8-12 nursing homes and, in some cases, their corporate organizations	11 nursing homes in 7 states implemented the intervention: developed/ used OCR forms on resident functioning/risk factors for pressure ulcers, incorporated timely report information, and began to use or explore technology options to sustain project activities Catholic Health Partners had 4 Ohio nursing homes participating in the project—provided a 'learning-lab' to examine how experiences of 4 facilities could serve as a model to standardize processes across an organization and to disseminate tools to other facilities

3. Project Evaluation and Outcomes/Results

The project's evaluation design involved the collection of baseline and follow-up data on (1) clinical measures (pressure ulcer incidence acquired in or out of the facility); (2) utilization measures (hospital admissions and ER visits); (3) operational measures, e.g., number of forms used prior to intervention; and (4) annual turnover rates and staff satisfaction measures. 

The combined average for 7 facilities that implemented project processes starting in April 2004 shows an overall reduction of 33% in the [CMS] quality measure (QM) of high-risk residents with pressure ulcer from pre-implementation to initial post-implementation time periods (through Quarter 3, 2005). Individual patterns for each facility show reduction in the pressure ulcer QM and percentage of high-risk residents with pressure ulcers.  Pressure ulcer prevalence in participating facility units dropped to about 8.7% on average, compared to the national average of 14%, which remained flat over the life of this project.  However, this may not be statistically significant because it is a small sample. Facilities that implemented the intervention more fully (e.g., regularly submitting forms, using the reports in regular care planning meetings) had better results—PU prevalence in the 5 to 6% range—than those that partially implemented the intervention.  

These early findings were updated with Quarter 4, 2005 data to summarize overall impact to date (by facility) on CMS QMs related to pressure ulcers.  It is important to note that the CMS QM for high-risk pressure ulcer includes in-house and externally acquired, as well as existing pressure ulcers, and is a measure for the entire facility. While this differs from the project's primary clinical outcome measure (in­house acquired pressure ulcers on participating units), the project team hypothesized that participating facilities focused improvement efforts on the unit(s) with highest risk residents; therefore, the interventions would impact the CMS QM for high-risk residents. Individual patterns for most facilities show reduction in the pressure ulcer QM percentage of high-risk residents. During Quarter 3, 2003, only two facilities were below the national average. For Quarter 4, 2005, six facilities were below the national average. All project facilities that have prevalence rates equal to or greater than the national average have decreased their prevalence from Quarter 3, 2003 by an average of 38%.

In addition to decreased pressure ulcer development, the project reduced the number of documentation forms that CNAs fill out at each facility, which reduces paperwork burden and provides more time for hands-on care to residents.  Information about residents is now available in "real-time"; quality improvement has shifted from reviewing data quarterly on a retrospective basis to using weekly clinical reports for timely resident care planning by all members of the care team. Communication among the care team reportedly has improved and collaboration across team members has increased.  Data needed for CMS and state survey reports are captured more easily and are readily available. 

4. Major Products

The workflow change process of using standardized documentation and timely feedback reports for improved care planning has been presented at many national conferences, including the 2004 and 2005 Annual Research Meetings of AcademyHealth, the Spring 2004 and 2005 AAHSA Future of Aging conferences, the 2005 AAHSA Annual Meeting, AHRQ's Translating Research into Practice meetings in July 2005 and 2006, and the Gerontological Society of America annual conferences in November 2005 and 2006.

5. Potential for Sustainability/Expansion after PFQ Grant Ends

Among the 11 facilities that participated in the project, four will not be involved in future spin-off projects, primarily because of turnover in the Directors of Nursing, who are key decision makers in nursing homes.  The remaining facilities are joining ISIS in a new Health Information Technology (HIT) project to continue the standardized documentation and reporting processes begun in this project; HIT is funded by AHRQ. 

Half of the participating facilities were part of larger systems or corporate chains.  This allowed corporate leaders to watch 'the experiment' and decide if it was worth adopting corporate wide. The Good Samaritan Society (GSS) was impressed enough to adopt the tools; according to the PI, 240 GSS facilities in 25 states are now using the same approach to documentation.  Mercy Health Partners, which had four facilities participating in the project, is rolling it out to more of their long-term care facilities. In addition, standardized comprehensive documentation by front-line staff, followed by timely reporting, has changed facility workflow. While designed around pressure ulcer prevention, it is applicable and helpful across clinical areas.  It is being used to facilitate improved resident care and better responsiveness to federal reporting requirements. 

Towards the end of the project's third year, ISIS had discussions with the Arizona QIO and initiated calls with QIOs in California, MD-VA-DC (Delmarva), Ohio, Texas, North Carolina, Idaho, Washington, and Rhode Island to explore their interest in replicating the model through the QIOs' nursing home quality improvement activities.  These discussions led ISIS to submit a separate contract proposal to launch this new approach to replication. AHRQ funded the contract, which began in September 2005. ISIS is working with California (Lumetra), Idaho (Qualis), Texas, Maryland (Delmarva), North Carolina, and Arizona QIOs.  The QIOs identified about 30 long-term care facilities; ISIS trains facility and QIO staff to help them implement the 'Real-Time' process using Digital Pen Systems or internal facility IT systems. 

In the final grant year, the project intensified its efforts to disseminate project activities to other long-term care facilities.  It will evaluate results and develop a plan for ongoing initiatives to continue expanding the number of participating sites, evidence-based medicine content, and data collection and reporting improvements.  To accomplish this, the ISIS project team is working in partnership with the AHRQ-funded contract to Delmarva Foundation for Medical Care, contract #290-04-0009, 'Real-Time Prevention of Pressure Ulcers,' which was funded in May 2006.  

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