With government incentives aimed at improving quality of care and outcomes, tools such as EMRs and electronic registries are increasingly critical. Integrating the two technologies is key to measuring quality and outcomes, but achieving that requires many pieces to fit together, from the technical to the practical.
Under the Affordable Care Act, physicians who do not participate in the government’s Physician Quality Reporting System (PQRS) or whose performance is not successful based on quality measures will incur a 1.5 percent Medicare reimbursement penalty beginning in 2015. Among the tools most critical to measuring quality of care are clinical data registries (CDRs) and EMRs.
“Registries are becoming very important in the management of patient populations and will be increasingly important over time,” says Richard Gliklich, MD, of Harvard Medical School and lead author of Registries for Evaluating Patient Outcomes: A User’s Guide, a publication created for the Agency for Healthcare Research & Quality.
Medicare-eligible providers not using functional EMRs with meaningful criteria will be penalized. In 2015, providers without appropriate EMRs in place will receive a 1 percent penalty, and the government is pushing for integration between CDRs and EMRs.
In a report released in December, the U.S. Government Accountability Office (GAO) found that CDRs are useful in analyzing outcome and treatment trends as well as measuring physician performance.
No easy technical solution
However, the GAO also determined that the Department of Health & Human Services is not doing enough to maximize the potential of CDRs to improve the quality and efficiency of care provided to Medicare beneficiaries. The GAO recommended facilitating integration of CDRs and EMRs.
First registry-based clinical trial A study evaluating the efficacy and feasibility of radial versus femoral access in women undergoing PCI was the first randomized clinical trial to be embedded into an existing registry. The Study of Access site For Enhancement of Percutaneous Coronary Intervention for Women (SAFE-PCI for Women) initially randomized 1,787 women to either transradial or transfemoral PCI. The rate of bleeding in the transradial PCI group was 2.9 percent vs. 1.2 percent, which resulted in the trial being stopped early for safety reasons. In addition to being open only to women, SAFE-PCI was also unique in its use of the CathPCI registry. Certain demographic information was electronically captured with no effort required by study administrators. Data specific to the trial were directly entered into electronic case report forms. The approach reduced the workload of the site coordinator by 65 percent per patient compared with traditional forms. It also may facilitate enrollment and lower costs. |
But experts who deal with the nuts and bolts of registry efforts say the process is no easy feat, complicated by technical, medical, legal and administrative challenges. “In most cases, the two do not talk to each other,” says Jeff Jacobs, MD, chair of the Society of Thoracic Surgeons’ (STS) Congenital Heart Surgery Database. “If they did, it would create opportunities to diminish the burden of managing data and increase the accuracy of both tools.”
The technical challenges, he explains, are to develop computer software that enables EMRs and CDRs to talk to each other and ideally, CDRs would automatically capture certain data from EMRs.
“Doing that well relies on a bunch of different factors,” says J. Brendan Mullen, BSFS, senior director for the PINNACLE registry, which is run by the American College of Cardiology (ACC) and collects outpatient data related to coronary artery disease, hypertension, heart failure and atrial fibrillation. “It relies on technical expertise to take data as they get recorded and transform them into a format that can be used in these registries.”
“Many people are trying to develop strategies to facilitate improved communication between EMRs and registries, but there are currently few successful examples of this happening,” says Jacobs.
Easing the burden
PINNACLE is one CDR that has some of that functionality. PINNACLE uses vendor software that enables some of the fields to be automatically populated from EMRs.
However, clinicians or other staff still must enter information into other fields as well, which is another major barrier to EMR/CDR integration. This data entry can be a significant burden.
“We need to take on more of the data entry burden at the ACC,” says Mullen. “We’ve developed technologies at ACC with our partners to go into data records and extract relevant information that clinicians and their teams already entered so we can get the data to do our research.”
Providers still need to enter the data, however, which can either cut into staff members’ patient care time or require the hiring of more data entry staff.
“Clinicians appropriately do not want to duplicate entry into two different electronic systems. Rather, they expect those systems to talk between themselves,” Gliklich says. The technology exists to enable the registry to appear within the EMR, even auto-completing many of the data fields from the EMR database, but these standards have not been put in place in most of the commercially available software products.
Entering data into some registries also can be cumbersome.
“I’ve had some physicians cynically comment that entering data on a single patient into the TVT [Transcatheter Valve Therapy] registry can sometimes take over twice as long as it does to actually perform an uncomplicated procedure itself,” says Ralph Brindis, MD, MPH, senior medical officer of external affairs at the National Cardiovascular Data Registry (NCDR). The TVT registry focuses on transcatheter aortic valve replacement and is run jointly by the ACC and the STS. It’s an even bigger problem for hospitals that participate in several registries to enter patient data into all of them, he adds.
Semantics & costs
Another challenge is terminology. “The terminology in EMRs is not necessarily harmonized with the terminology in a clinical registry,” Jacobs explains. “There are ongoing efforts to harmonize definitions of basic terms like heart attack and stroke, and even the definition of death can vary from one database to another.”
Gliklich credits the American Heart Association and the ACC for their efforts to standardize registry definitions, but some terms lack common meanings. Various specialties also need to standardize what data need to be collected.
“An EMR has basic medical record information in it, but clinician groups need to determine what they think should be standard for certain conditions,” Gliklich says. Without that data standardization, software vendors are unable to create software that can better integrate EMRs and registries on a broad scale.
Providers also must heed the need for patient privacy, Jacobs says. Patient privacy must be protected whenever patient data leave an EMR en route to a registry. Providers also need to ensure that research participants give informed consent.
The experts argue that investing in registry participation will help improve the quality and efficiency of healthcare. The investment can be costly, they explain, depending on the registry.
Recent successes include the use of the CathPCI registry to capture data in the SAFE-PCI for Women trial (see sidebar) and use of PINNACLE to recruit participants for clinical trials.
Brindis says participating in the CathPCI registry costs $5,000 per year, not including the costs for staff to enter data or the time and expense of making sure data entered are accurate and correct. Partnering with a software vendor also can add substantial costs that vary depending on the services a hospital contracts to receive from the software vendor.
“Depending on the size of the hospital and the number of registries they’re in and the number of procedures they perform, registry participation can be well over a $100,000 investment,” he says.
Despite the costs, more and more providers are choosing to participate because of financial incentives and to have the ability to measure quality of care and manage populations of patients as well as support research initiatives. In fact, Jacobs says most physicians participate in heart surgery registries.
Look to societies
In the registry evaluation handbook, Gliklich and colleagues say that interoperability between EMRs and registries should be standard enough to allow any provider to interact with a registry without needing customization or permission from the software vendor.
“For many providers, the most accessible registries will be those being developed by their specialty societies,” Gliklich says. “In choosing an EMR, providers should find out if the EMRs they are interested in are integrated with their society’s registry.”
A newer tool, called either retrieve form for data capture or the Healthcare Information Technology Standards Panel’s HITSP-TP50, will become more widely available. “If the EMR has that in the product, that’s the best case as any web registry could integrate with it,” Gliklich adds.
Although the integration process is improving, full integration is not likely for some time. “The resources, expertise, tools and costs are substantial,” Brindis says. “The need for full integration is great but attaining this goal involves incredible complexities.”
He adds that professional societies, vendors and the government are working toward making integration a reality.
“We are still, however, a few years away from true success,” he says.