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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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: https://info.ahrq.gov. Let us know the nature of the problem, the Web address of what you want, and your contact information.
Please go to www.ahrq.gov for current information.
AQA Invitational Meeting
Status Report on Quality Measurement
Elizabeth McGlynn, RAND Health
Elizabeth McGlynn opened the meeting by highlighting the advances that have been made in measuring quality and reporting on quality. These include good tools and a wealth of information. At the same time, she stressed that the available information is often difficult to find and interpret and said that much of it offers different answers to the same questions. McGlynn also said that American adults receive only about one-half of the recommended care, and that number drops considerably when it comes to care for geriatric conditions.
McGlynn said the question for this meeting is, "How do we get from where we are to where we want to go?" She said the "next frontier" for quality measurement and reporting is at the physician practice level (either individually or in larger groups).
In order to frame the discussion, McGlynn reviewed the work of the Strategic Framework Board (SFB), which was formed in December 1999 to develop a strategy for national quality measurement and reporting. The aim of the board's work is to improve the health of the American public through the use of quality measurements. The SFB is also charged with articulating the guiding principles and priorities for a national system and identifying potential barriers to successful implementation of any proposed solution. McGlynn discussed the SFB's conceptual framework for a quality measurement and reporting system, saying that the aim is to:
- Evaluate the degree to which the U.S. health care system is providing safe, effective, timely, and patient-centered care.
- Assess whether the delivery of high-quality health care is efficient and equitable.
- Enable substantial progress to be made toward achieving established national goals.
- Provide easily accessible information on quality to a variety of audiences (including consumers, purchasers, health plans, and providers) in order to facilitate individual and collective decisionmaking.
- Provide information that regulators, purchasers, health plans, and providers can use to support continued improvement and the achievement of goals.
McGlynn said the system must be evidence-based and that in order to have an evidence-based system, you must have data. She said a major product of a national quality measurement and reporting system must be evidence about quality. She noted that it was also necessary to develop strategies to take action when evidence is weak or absent, and to routinely identify gaps in evidence and develop strategies to fill in the gaps. Overall, stressed McGlynn, the challenge is to find a balance between absolute certainty and the need to treat patients on a day-to-day basis.
The SFB has advocated a simultaneous top-down (goal setting) and bottom-up (front-line care delivery) approach to selecting measures and reporting strategies. McGlynn said that common measures were the currency of exchange that links what happens on the front lines of care with efforts to achieve national goals for quality improvement. Effective reports, she said, will "provide the right information, in the right form, at the right time, for the right people to make the decisions that are right for them."
McGlynn stressed that it was important to recognize the tension between national goals (which provide a focal point for leadership and action) and local concerns. She cautioned that while the choice of these national goals would shape the choice of quality measures and reports, adopting goals is ultimately a local issue. Goals require buy-in if they are to be adopted and result in quality improvements.
A strategic plan for adoption is needed, she said, and it must include public awareness, an improved product, and systems that reward high-quality providers. In addition, although leadership at all levels of the system is essential, local advocates and channels for connecting participants must also be developed. McGlynn added that without a strategic plan for adoption, quality measurement and reporting as a tool to drive improvement may not realize its potential.
The purpose of a national quality measurement and reporting system, stressed McGlynn, is to improve quality. She noted that:
- Transparency is the goal of information generated for accountability and selection.
- Change in care delivery is the goal of information generated for improvement.
McGlynn added that everyone's skills—those of consumers, patients, providers, and purchasers—need to be improved in order to use this new information effectively.
We are at a critical juncture, said McGlynn, with a national focus on expanding the use of health information technology. Health information technology is a huge tool, but we need to figure out how to use it effectively. In addition, McGlynn said that while there were multiple experiments under way to reward providers for reporting and for improved performance, it is important that people not draw conclusions too quickly. She added that while there were very exciting activities under way, including engagement by specialty societies in a dialogue around physician-level measurement, there is currently no strategic plan to coordinate these activities.
The SFB contends that the key dimensions of a strategy must include:
- Selecting national goals.
- Conducting a leverage analysis.
- Choosing measures.
- Reporting results.
- Using the information to design a data system and the infrastructure to allow these activities to take place.
Regarding national goals, McGlynn said that the Institute of Medicine priority areas represent a starting point for discussion. The issue, she said, is whether the presence of national goals (or the process of developing them) would facilitate the engagement of individual physicians.
McGlynn took meeting participants through some examples of what a national goal might look like. These included such objectives as "improve the quality of life for persons with depression" and "reduce mortality from heart disease" as well as "improve the coordination of care for patients with multiple health care needs" and "ensure the safe and effective use of medications."
The political reality, said McGlynn, is that no activity will occur unless there is a centralized funding mechanism. Operationally, she said, while it is possible to develop a starter set of goals, testing them in multiple clinical areas across the continuum of care will be necessary to evaluate the credibility of the information technology and to assess the credibility of pay-for-performance measures. McGlynn also acknowledged that agreeing on a starter set of goals for clinical areas is likely to be easier than developing a mechanism for updating them and for ongoing revision.
To develop national goals, it is essential to conduct leverage analyses. This is useful for determining what set of actions is most likely to make a difference and where the locus of control for action resides, said McGlynn. She took meeting participants through the steps of a leverage analysis for the goal of reducing mortality from heart disease. The example outlined the steps that take place after a heart attack (from on-the-scene cardiopulmonary resuscitation [CPR] to the adequacy of the trauma system), as well as the events that precede a heart attack. The leverage opportunities are primary and secondary prevention—because the best way to survive a heart attack is not to have one in the first place.
Next, McGlynn said that under the SFB "rational" design, measures will flow from the leverage analysis. She said the measures designed for hospitals and health plans likely cannot also be used at the individual physician level.
The SFB has recommended integrating performance measurement for accountability and for improvement. There are two approaches:
- An internal measurement model.
- A transparency model.
The option you select will drive the health information system you create. The first model argues that data should be used internally to measure improvement and guide care and assumes that individual physicians aren't really in a position to make major quality improvements. Under the transparency model, if people knew how well a hospital, health plan, or physician were performing, said McGlynn, they would use this information to make choices about providers. In other words, they would make rational decisions on where to go for care. McGlynn said that the SFB tried to make the point that both approaches can operate from the same set of data.
Looking again at the political realities, McGlynn noted that beliefs about the availability of data (now and in the future) drive perspectives on measurement opportunities. She noted that there has been considerable debate about whether measures should be proprietary or in the public domain. In addition, she pointed out the tension between those urging deliberateness and those urging speed and said that it was important to find a balance between the two. McGlynn added that there was a lot of work to be done to align measures to multiple uses.
Turning to data aggregation and reporting, McGlynn stressed that for data to be used for accountability, pay for performance, quality improvement, and other decisionmaking, they must be assembled, processed, packaged, and delivered to multiple sources. They must also be collected from multiple sources. If not, we will end up with suboptimal or potentially misleading information, she warned.
Collecting good data, McGlynn noted, will not be easy. Not only do physicians see multiple patients, but patients see multiple physicians and physicians have multiple contracts. Coordinating the information from these multiple sources will be a challenge.
The political and operational reality, said McGlynn, is that data are power. This means you "have to give to get," she said. McGlynn said that some pooling of data—especially at the physician level—will be necessary, and that this raises a number of challenges and concerns:
- What data elements will be pooled?
- Who holds the pooled data?
- Is that person trustworthy and competent?
- Who decides how the data will be used, and will there be opportunities for review prior to the release of those data to health plans, purchasers, and the public?
- How does this get paid for?
Prior successes offer some lessons, said McGlynn. She highlighted several:
- Those being measured have to be at the table and have a clear role in policymaking.
- Starting where there is common ground allows for more rapid progress.
- A trusted, neutral broker is essential.
- Initial efforts probably will not please anyone.
- Progress will be too slow for some, too fast for others.
- The availability and nature of funding will drive the process.
- People, not technology, will need to solve the critical problems.
Finally, McGlynn noted that while there are a lot of lofty visions about where we would like to go, we need to figure out how to get there from where we are today.
During the discussion that followed, there was a followup question on trust and competence. Elizabeth McGlynn said that trust depends upon each person's perspective, and that it would be necessary to find data brokers who are trusted across diverse segments of the health care community. She added that the competence issue has to do with processing the data—and that even good companies can mess that up from time to time.
So who holds the pooled data? McGlynn said that, in her opinion, this was not something that should be handled by a single entity. At the same time, she reiterated that there is a lot of waste now in how data are collected, and that the lack of data pooling and data sharing is placing a huge burden on the system.
Another participant noted the need to intersect the discussion on quality with a conversation about the data. One way to get started, suggested McGlynn, was to build the ideal Web site. Depending on your perspective (physician, provider, hospital, and so forth), what information would you like at your fingertips?
Another participant suggested that there was a missing element in the discussion to date: technical assistance to help physicians implement health information technology.
There was also discussion about whether structures can be "imposed." While one person noted the need for a multi-stakeholder forum, another said there were structures in place already and questioned whether a larger group could effectively be put in place. One participant cautioned that all the stakeholders must be at the table if viable solutions are to be achieved. There was discussion about whether the National Quality Forum (NQF) is the appropriate vehicle for this process or whether a different group could accomplish more. One person observed that the people represented in the room were "not natural bedfellows" and said that the real question wasn't who would undertake these activities but whether everyone could commit to focus on figuring out how to measure quality, report on it, and advance the public health.
We need to think about the different structures in place and how we do business, said one participant, so that physicians feel these are credible measures (versus merely being imposed on them). Another person suggested focusing on efforts already under way and identifying what each stakeholder can do to advance common objectives.
While the NQF brings together stakeholders to discuss what to measure, said another participant, the harder question is how to translate measures into day-to-day patient quality of care. This was followed by a comment about how to get agreement on a handful of measures (given the large number of measures out there). McGlynn said that, in her opinion, the issue was not the number of measures, but rather having an information technology system that allows ease of data collection and analysis. She compared this to an airline black box, which collects information from multiple sources.