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 It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work.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
Creating a Strategy for Data Aggregation and Stewardship
David Kibbe, American Academy of Family Physicians
David Kibbe said that the country was long overdue for a national strategy for data collection, aggregation, measurement, and reporting on quality processes and outcomes for common expensive, chronic illnesses. If the next frontier is data collected from the physician and practice levels, he said, then we must address barriers and opportunities that exist at the physician and practice levels. These include trust and competency; adoption of technology; and new models for data exchange (and new skills for managing health information) that include the patient as part of the data supply chain.
Regarding trust and competency, Kibbe said that if data are to be collected from physicians, then data stewardship must be practiced with complete transparency. He added that issues of data ownership and rights of use must be addressed, and he warned against data arbitrage.
Regarding the adoption of technology, Kibbe said that performance and quality data must become the routine by-product of the delivery of care using an electronic health records (EHR) system. It is too expensive to get the data any other way, he said. Kibbe also said that certification of electronic health records, which includes data collection and submission standard features, must be completed nationally. In practices that do not have electronic health recordkeeping systems in place, he said, technology must lead in the direction of electronic health records along pathways that are not dead ends (so people don't have to throw away their existing systems and start all over).
We are further along with electronic health recordkeeping in physician practices than many people assume, said Kibbe. He stressed that these early adapters are not innovators, nor do they care about the technology—but they want it to help their practices survive and thrive. The return on investment is becoming clearer every day, he said.
Kibbe noted that there were several key barriers to the deployment of a national electronic health recordkeeping system. These include:
- High prices (solution is affordability and price transparency).
- Risk of implementation failure (solution is evidence-based technical assistance).
- Lack of connectivity and interoperability (solution is for EHRs to be the "central nervous system" of the practice).
- Confusion about the product and company reliability (solution is EHR product certification, which is well under way).
- Wide variability in contracts and business practices (solution is standard contract language and clarity about business practices after sale).
- Access to capital (solution is low-interest loans).
Regarding new models for data exchange, Kibbe noted that new and more efficient models of data exchange are required, along with new skills for managing them. It's not just about data producers and users, he said. Rather, he said, we need to have a health data supply chain (and a market for these data) that involves almost everyone. Among these are physicians, patients, hospitals, emergency departments, long-term care facilities, and researchers. We need to recognize that there is a market for data that have multiple uses, are recycled, and pay for themselves, said Kibbe.
We can stage a national data submission implementation standard, said Kibbe. To do this first requires agreement—which we have—that standard measures are essential. Now we need to develop a blueprint for electronic health record certification and find the capital to enable national deployment of health information technology in small medical practices. And, of course, we need data, he added. Kibbe said that what was needed was a uniform architecture for data collection, management, analysis, and reporting that takes into account the complexity of the data supply chain.
Kibbe stressed that data are just data (and thus that people should not make a distinction between administrative and claims data). The question, he said, is how to use data from multiple sources. He added that the continuity-of-care record is a basket, and that there was no reason why one particular entity should be responsible for populating the whole thing. Instead, it would make sense to have multiple entities filling it out and providing useful information, Kibbe said.
Finally, Kibbe stressed that a key question that remains is how to engage physicians. When they complain about accuracy, he said, then tell them we'll work on it to improve the data reporting. As long as physicians are engaged, then we can make big strides, said Kibbe. Right now, he said, they don't see much in this for them.
The discussion started with a comment about the "artificial dichotomy" between claims and administrative data. The data that are critical to care come from all over the place, including other physicians, said one participant. We have to be very careful about how we think about using the data. If we are trying to judge the competency of a physician, he said, that's one thing. But if we are interested in the functioning of the physician within the system he or she is imbedded in, then we want to know whether a test that was ordered was done, and that's a different usage.
The participant added that we need research, thinking, and demonstration projects on how we record clinical data in the office. While he liked the idea of electronic medical records, we need to rethink how we record data because we are, over time, losing a lot of the critical clinical information we receive.
In response, Kibbe noted that there has been a tendency to transfer the paper documentation model to a computerized system. He said that the continuity-of-care model has thrown a wrench into this approach because the summary health information is built differently. There's a real struggle going on that will have profound implications on how we measure quality in the future, he said.
One participant noted that the health care community is in the process of transitioning to electronic health records, and noted that health plans have much more administrative data available than do individual doctors. A second person echoed the latter comment and asked whether the data available through the health plans could be accessed to inform physicians and patients. Someone else suggested that it was necessary to rethink the whole issue of data ownership.
Another participant stressed that the health care community should not wait until there is a critical mass of electronic medical records before doing something with the data. Instead, we should talk about using administrative data so we can get started by January 2006 with some degree of initial credibility.
There was a question for Kibbe about what he thought American Academy of Family Physicians (AAFP) members wanted to see done next. Also, what would make AAFP members comfortable regarding data stewardship?
In response, Kibbe said his members would like to get rid of all the paper, and they would like to make their practices financially successful. But they aren't really all that interested in submitting quality data, he cautioned. To the extent that we can provide reasonable, financial incentives to submit these data, he said, I think we could get considerable interest.
There was a followup question about how to make it "worth it" to providers to submit data. The participant noted that she had counted 86 existing incentives programs, and that most are considered poor. If these incentives aren't good, then what would be an attractive motivation? she asked. In response, Kibbe said the first step was to provide help in purchasing and implementing electronic data recording systems. Once a certified system is up, he said, then the physician can submit data (and become engaged, and get extra money to do so). Next, he said, provide additional incentives for physicians to reach their quality-of-care targets.
One person asked about trust, and how to structure the custodial function (vis-à-vis the data) to give everyone a much-needed level of comfort. Another person observed that gaining trust was going to be difficult. Yet another focused on credibility. Any time you hide things, even if for a good reason, it causes concern (i.e., even though we expect the public will find a composite measure most useful, consumers should be able to click down and locate the details), said the participant. The person added that there was also a difference between disclosing measures and disclosing all the data in the database (i.e., multiple stakeholders may agree on the measure, but perhaps not all will agree on disclosing all the data so someone else can put it together as they see fit).
It sounds like we need short-, medium-, and long-term data aggregation strategies, said one participant. There is a lot going on, and the question is whether we can light a fire and find the two or three models that can be applied everywhere in the country.
A lot of the focus is on the Medicare population, noted one participant, who said his focus was on chronic diseases (versus, for example, employers, who are likely to prioritize different measures). The person noted that progress on the health information technology front is coming quickly, because both plans and payers have committed to help fund adoption of electronic health technology. But the question is: What do we get if we help fund it? (Quality data, replied someone else.)
One participant noted that the American College of Surgeons (ACS) is working with the Department of Veterans Affairs on Web-based data collection and management. With AHRQ support, ACS is in the process of making the program available to any hospital (both VA and other) that wants to use it. The system is totally secure (because the data are encrypted), and data are only given to the hospitals that report data, he said. The data include preoperative, postoperative, and operative data. Another person noted that ACS was one of several groups that have maintained databases, adding that the real issue is one of data aggregation. How many of these separate systems will be mainstreamed? he asked.
In response, another participant noted that both access to data and public disclosure are critical. She urged everyone to think about multiple uses. "Collect once and use many different ways," she said.
One person raised the related issue of creating composite measures and massaging data for different uses. I need high-level aggregate data for some stuff, said the person, and I need to package data in different ways. He added that it was important to pay attention to how data, after they have been aggregated and analyzed, are further aggregated and manipulated. Most consumers, he said, don't want to look at 25 measures when choosing a physician.
Another participant suggested that this discussion was really about two separate things: building roads (the infrastructure) and building houses. We need to recognize that each needs to be funded, he said. First we need to get there (data aggregation), then we need to address what we are going to do with the information once we have it.
Finally, there was a remark about the need for synergy at each level, and the need for "vertical parsimony" (so that we have the same measures for reporting, quality, and national tracking). This is easier said than done, cautioned the participant, because the availability of data is different at different levels.