Network seeks uniformity in data standards
Network seeks uniformity in data standards
Universal definitions is ultimate goal
The National Institutes of Health (NIH) Roadmap for Research seeks many improvements, and one of these is to make research more efficient.
As a part of this goal, NIH sought proposals from clinical research networks, which are working together to make data standards uniform and translatable.
"Within our clinical trials group at Duke University, we have a number of investigative networks, some more formal than others," says Robert A. Harrington, MD, professor of medicine in the division of cardiology at Duke University Medical Center in Durham, NC.
"What we're best known for is a cardiovascular trials network, which is an informal collaboration of several hundred practices throughout the United States who regularly participate in our clinical trials," Harrington says.
"We act as a coordinating center and coordinate the activity," he adds. "So we put in a proposal to the NIH, structured around a group of investigative sites, creating, implementing, and sharing best practices for clinical trial networks."
The group aims to provide tools that research sites can use to do a better job of research, Harrington says.
These include tools to improve communication, operations, education, training, budgeting, and good clinical practice in clinical trials.
"The second aim dealt with the concept of inefficiencies in research, due to redundancy in data collection," Harrington says.
"You can have a patient come in and fill out a medical record, and if the person is involved in a clinical trial, he will fill out several forms, and then fill out more forms for billing," Harrington explains. "Inherent in the clinical research process is an inefficiency in how data moves."
So the second goal resulted in creating a pilot project to coordinate a group of stakeholders in cardiovascular medicine and the research community in the creation of a public standards forum, with a set of data element standards regarding acute ischemic heart disease, he says.
"Aim three was how does a network identify and develop research ideas, and aim four was to build a model of information technology platform/tools for a network," Harrington says.
The project, which was funded for 1.5 years, is underway. Here's how it works:
• Collaboration: All organizations in the medical and research community that have a stake in cardiovascular medicine collaborate and participate in the creation of data standards.
The premise is that if everyone defines terms the same way then the data may be usable in many formats, rather than having to be created separately for different needs, Harrington explains.
"If you have a heart attack and come into the hospital, going through the emergency room and cath lab, etc., then all of your information should be collected in a way that allows data from the medical record to be transported into the clinical trials structure and also transported to the hospital billing system, etc.," Harrington says.
"We all have to agree upon what is the terminology for medical issues and technical standards to allow the transfer of that data," he says. "This started out as a network need, and we have built this into a project that is much broader than our network, but which has direct relevance to people working collaboratively with us."
• Gathering information: Harrington always knew data collection and transfer was an issue, but he didn't know the magnitude of the problem until he began working on the project.
"We've spent a fair bit of time gathering information," Harrington says.
All major cardiology groups in the United States are involved, including the American College of Cardiology, the Society of Thoracic Surgeons, and the American Heart Association.
"We also gather people interested in cardiovascular disease from a research perspective," he says.
Federal agencies and the VA hospital system have been involved, as have stakeholders from the medical products community, including device, pharmaceutical, and software companies, who each have some interest in standardization, Harrington says.
"C-Disc, a data standards group, has been coordinating this effort in collaboration with us," Harrington says. "They are a not-for-profit public standards group that has credibility in the standards creation world, but has largely been dealing with data standards from a technical perspective."
All of these stakeholders were asked for examples of their data elements in this disease state, and these examples have been organized into databases, Harrington says.
The databases offer the opportunity to look across the different elements for different definitions, Harrington notes.
• Selecting standard data elements: "We're now in the process of winnowing these data elements down, and ultimately, we'll be engaged in the vetting process of deciding which elements to include," Harrington says. "Ultimately, we'll reach consensus on which data elements should be standard."
The key will be to get all of the stakeholders to agree upon a definition set that may ultimately influence how ICD-9 codes are constructed, Harrington says.
"You'd like to move data seamlessly from one area to the next," he says.
Once all of the data definitions are agreed upon, then the collaborative group could do a project in which one or more centers will take data entered into the hospital system for clinical care and demonstrate how using common definitions and transport standards will enable the data to be taken out and used for other purposes, Harrington explains.
"With one set of data, we'll show it can be used for research purposes, quality improvement purposes, financial reporting, etc.," Harrington says. "It's a methodical project to show how one creates data elements; what the issues are along the way, and how successful that project might be in a very narrow scope."
The NIH is working toward making these efficiencies a national goal, he notes.
"Re-entering data creates errors, and that's where a lot of inefficiencies come from," Harrington says. "People would like to participate in research, if it were easy."
• Using a single set of data elements: "Can you imagine how much easier recruiting for research would be?" Harrington says of a standard data set. "The President is calling for electronic health records for all Americans in the next 10 years, so the American health system clearly is moving into the inter-operability phase."
Once there is an agreed-upon set of standards, data could transfer very rapidly from one city's health care system to another, as well as from clinical care to clinical trials, he says.
"Imagine having a doctor see a patient with a certain disease and having all data entered into an electronic health record," Harrington says. "At the end of the visit, it's unclear about what is the best treatment, and so right then the doctor could ask the patient to participate in a clinical trial."
The patient would be randomized in the trial, and the data would be partitioned out for pieces that are important to clinical trial research, Harrington says.
"If that doctor is participating in the trial, he or she could basically signal electronically that this patient is in the clinical trial and the data would move from the clinical health record to the clinical trial case report form," Harrington says. "This would be the ultimate goal; people would get the benefit of participating in research without the duplication, which makes the process inefficient."
• Looking at best case scenario: There might be multiple research networks that overlap with one another, but all are assisted by Web-based technology, Harrington says.
There would be a platform on the Web to provide the networks with information about research and how to get involved with protocols, including information about how to train staff, enroll patients, and manage clinical trial operations, he adds.
"This would allow more practitioners to participate in research, which is certainly a goal of the Roadmap," Harrington says. "It's in the best interest of the American health system because we'd be able to get the answers to more questions, more quickly."
The fragmented system, as it stands today, leaves gaps in medical knowledge, since there are a lot of things happening in medical practice that people are not studying, Harrington says.
"You can envision a future where questions are asked about whether anyone is doing a particular practice, and this would help link them to what is being done," Harrington explains. "It would take down barriers of participation for both patients and clinicians."
With common data elements, it would be possible to pull various investigative sites together in a network in which they use the same definitions and data, creating a community of investigators, Harrington says.
"If you have people believing that being part of the community of investigators is important, then technology is not the issue," Harrington says. "For instance, you have MySpace connecting hundreds of thousands of kids."
Harrington says he came up with the MySpace analogy when his teenage son said to him, "Dad, what you're describing sounds like MySpace for researchers."
Another analogy is Wal-Mart's information data tracking system, which collects detailed information down to each item sold in each store around the world, he says.
"It would seem to me that human health is a little more important than what toaster you buy, yet nobody's done it for the health care system," Harrington says.
[Editor's note: For more information about the data element standards project, you can contact Robert Harrington, MD, at [email protected].]
The National Institutes of Health (NIH) Roadmap for Research seeks many improvements, and one of these is to make research more efficient.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.