AQA Invitational Meeting Summary

Pilot Projects and Value Exchanges

Carolyn Clancy, Agency for Healthcare Research and Quality (AHRQ)
Peter Lee, Pacific Business Group on Health

Carolyn Clancy opened the discussion on the pilot program (known as Better Quality Information for Medicare Beneficiaries, or BQIMBs) by noting that all six subcontractors have now received the Medicare data. She indicated that the subcontractors were now working to aggregate the data, and the first reports of aggregated data at the practice level (and the detailed issues that need to be addressed) are due in July. Once the data has been aggregated, Clancy said, then the subcontractors can work together to address data attribution rules. The aim is to report on data at the individual physician level in November.

Peter Lee added that the BQIMBs are the starting point for the process, as they are only looking at a handful of measures within the initial AQA starter set. The objective, he stressed, is to aggregate the data. He termed the value exchanges "the next generation of pilots," and said the objective of these pilot projects would be to use the data in order to begin to give consumers, payers, and others good information to make decisions.

Lee noted that there was clearly a lot to learn about how to present the information in ways that are meaningful for consumers and that give clinicians good information to drive quality of care. He said the first step to becoming a value exchange (under the Department of Health and Human Services' Value-Driven Health Care Initiative) is for a organization composed of multiple stakeholders to define itself as a community leader. Lee said that these communities would then have access to data.

Finally, Lee pointed out that it was not yet clear how much of this work can, should, and ought to be done locally, as opposed to being reported and aggregated to a central data source. He expressed concern about having 70 separate communities using data differently-and thus voiced support for developing a nationally consistent platform.

Discussion

Opening the discussion, one participant asked what benefit a community would receive from becoming a value exchange. In response, Lee said that the learning network extends beyond merely a LISTSERV® and that AHRQ has devoted resources to help communities learn to get it right. Lee added that value exchanges will get the results (but not the data—although they may subsequently get both data and financial resources). You have to be a value exchange before you ever can hope to get data and dollars, he said.

Clancy indicated that, beyond the pilot sites, there were only about 14 communities that currently have robust, multistakeholder collaborative processes designed to assess quality of care. At the same time, she stressed that health care is a local event, and that fixing the system can be done only by those providing the care.

Lee noted that AHRQ had issued a recent notice in the Federal Register seeking public comments on the plan to establish value exchanges.

A participant asked whether the goal was to harmonize the multiple efforts. In response, Clancy said that it would be difficult to do so and maintain AHRQ's hallmark of transparency. At the same time, she said that figuring out how to aggregate data, post it, and maintain transparency was part of the interesting conversation that lay ahead in order to improve the quality of care.

Another participant observed that the value exchanges appear to be akin to Regional Health Information Organizations (RHIOs). He also expressed concern about how to get reports to physicians that can be acted upon when the value exchanges start to receive results (and possibly data as well). More immediately, he asked where the data would be aggregated for the value exchanges. Finally, he suggested that the AQA submit comments on the value exchanges to AHRQ.

Clancy replied that AHRQ was still working through the question of where the data would be aggregated. Regarding the suggestion to submit comments, Clancy said that the recommendation would be referred to the Data Sharing and Aggregation Workgroup.

In response to a question about the model for value exchanges, Clancy said that what was clear is that an entity must demonstrate a track record and/or a very clear commitment to engaging multiple stakeholders who are in the business of sitting down and assessing quality of care. Lee added that it was not clear, beyond acting as conveners, what role the value exchanges would play.

Next, a participant observed that it seemed clear that the pathway to Medicare data was through the value exchanges. She said this could be a challenge for national health plans with employees in multiple (or all) States. She asked whether the plans would need to participate in every value exchange, or whether perhaps there will be one value exchange for each geographic area. Clancy replied that the solution had not yet been determined.

In response to a separate question about what the data from the central aggregator will consist of, Clancy replied that the reports would derive from data aggregation from both private and public sources. She said the question being considered right now was whether the public and private aggregators would be different.

Another participant asked if the private plans would be sent a request to contribute their data. Clancy replied that it was quite possible and added that data requests would be for specific data based on performance reports.

Another participant said he was interested to see the relationship between the National Health Data Stewardship Entity and the value exchanges. In response, Clancy said that she saw a strong interaction between the two. She added that the value exchanges were trying to find a way to create virtual networks within communities.

One participant expressed two "overwhelming problems" with measurement: (1) sample size; and (2) the fact that putting large amounts of data together won't solve the problem, because not all data are equal. He noted, for example, that some health plans have superb pharmacy data while others may have very reliable lab claims across their populations. But when you mix these data together, he said, you find the gaps. He wondered how many AQA- or NQF-endorsed measures would be accurate and reliable down to the individual physician level.

Another participant asked about the difference between the value exchanges and the pilot projects. In response, Clancy noted that the value exchanges could potentially look at a broader array of measures approved by the AQA and the National Quality Forum. Lee reiterated that the BQIMBs were designed to address how to do aggregation, collection, and reporting (i.e., about getting the right data so it is usable), while the value exchanges are designed to look at how to use the data. The value exchanges are about use, he stressed, so the value exchanges need the data to be usable and the rules of the road to be predetermined.

Finally, a participant suggested that "elevator speech documents," very short questions-and-answers, on the BQIMBs and the value exchanges be developed and posted on the AQA Web site.

Closing Remarks

Wrapping up the meeting, Carolyn Clancy noted that the schedule for each workgroup's upcoming conference calls was posted on the AQA Web site. She then thanked attendees for their participation in a very helpful discussion and adjourned the meeting.

The next AQA meeting is scheduled for October 18, 2007.

Current as of July 2007

---

Previous Section        Contents                        

---

Internet Citation:

AQA Invitational Meeting. Summary. July 2007. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/qual/performance7/

---