Feature Strength in Numbers: The Power and Potential of Clinical Data Registries More than 110 qualified clinical data registries are helping improve health care outcomes while demonstrating the value of specific interventions. Here's a look at some of them, and the benefits that APTA's Physical Therapy Outcomes Registry can offer the profession. By Michelle Vanderhoff | November 2017 The transition to value-based care may seem daunting to some physical therapy practices. It requires accountability for patient outcomes, evaluating the effectiveness of interventions, and making technical and organizational changes—all within an evolving and sometimes complicated payment system. The ability to demonstrate and quantify the value of physical therapist (PT) interventions is powerful leverage, however, when it comes to fair payment for physical therapist services currently covered under insurance plans, as well as for some services not covered, such as prevention. Data analytics tools—and particularly clinical data registries—can help providers meet these challenges in their own clinics and position the profession for success within the health care ecosystem, say PTs and others interviewed for this article. This includes APTA's Physical Therapy Outcomes Registry, which collects and aggregates electronic health record (EHR) data to help PTs make well-informed clinical decisions and track and benchmark clinical outcomes. Since payers don't have consistent outcomes data to review, explains Heather Smith, PT, MPH, they control physical therapy expenditures "through blunt cost-reduction measures" such as multiple procedure payment reduction (MPPR) and the therapy cap on the Medicare side, and utilization review on the private payer side. And even if payers did look at outcomes, she observes, they would find it difficult to evaluate them because "we don't have a set of uniform measures that are applied across all therapists." Smith is APTA's director of quality. Private payers, she notes, are "definitely interested" in outcomes data. Understandably, part of the reason is to enhance their own bottom line. Additionally, though, pressure is coming from the group coverage market as employers attempt to curb costs. Almost 40% of large employers are starting to incorporate value-based care into their health benefit plans, according to the National Business Group on Health.1 Clinical registries can help immensely with quality improvement, although clinicians must balance the benefits of participation with the burden. More than a decade ago, some physician specialty societies began to use clinical registries to measure the effectiveness of care. Back then, the registries were manual-entry databases. Today's technology enables many registries to pull clinical data directly from EHRs. This has been a boon for the American Academy of Ophthalmology (AAO) IRIS® Registry (for Intelligent Research in Sight). IRIS launched in 2014 and already includes 16,700 participants and more than 148 million patient visits. AAO Medical Director of Health Policy William Rich, MD, who oversees the IRIS Registry, previously was part of a private practice. Rich reports that his colleagues' hesitations dissipated when they realized how little additional work the registry would require. Producing "Revolutionary" Findings While it's taken some time to build up a sufficiently robust amount of data to draw deep insights, the IRIS Registry already has yielded some surprising findings. For one eye condition, 90% of ophthalmologists had shifted from an older procedure to a newer one Rich calls "high tech" and "elegant." But an analysis of more than 100,000 patient cases revealed that the 90-day re-operation rate for the newer procedure was 2.5 times greater than the older one. The discovery amazed Rich, who also is past president of AAO. "That's revolutionary," he says. "It's going to change the way we practice." This reaction is echoed by Frederick Masoudi, MD, MSPH, chief science officer of the American College of Cardiology's (ACC) National Clinical Data Registries (NCDR). After he and his colleagues assessed registry data, they found that many patients receiving an implantable cardioverter defibrillator (ICD) for primary prevention, and without a clinical indication for pacemaker function, were receiving dual-lead defibrillators, when a less-expensive, less-risky single-lead defibrillator would suffice. NCDR consists of 10 clinical registries focused on cardiovascular procedures and conditions. Masoudi's team introduced a quality metric into the registries to prompt health systems and physicians to "think more about which device is best for their patients." Researchers have published hundreds of papers based on registry data, and their findings have changed clinical practice guidelines, according to Masoudi. In total, NCDR includes more than 100 million patient records from 15,000 cardiovascular specialists, with approximately 3,000 facilities participating.2 A Range of Registries There are several different kinds of registries. Some, such as NCDR's LAAO Registry (for Left Atrial Appendage Occlusion), support both Food and Drug Administration (FDA) post-market surveillance studies and CMS coverage-with-evidence decisions (in which CMS requires contribution of data to generate clinical evidence in exchange for reimbursement). Some are outpatient registries focused on treatment of a specific condition or disease. Others are specialty-specific registries, such as IRIS and the Physical Therapy Outcomes Registry. Many society-sponsored registries, including APTA's, are Qualified Clinical Data Registries approved to submit Merit-based Incentive Payment System (MIPS) quality data to the Centers for Medicare and Medicaid Services (CMS), which can be helpful for practices whose EHRs are not equipped to submit quality measures. While some EHR vendors are beginning to create their own "registries," they are "hampered by lack of data standardization," Masoudi explains. Others provide "quality feedback" based on claims data, but don't offer the granular clinical detail that registries from specialty societies do. Two things distinguish specialty society registries from EHRs, according to Masoudi: (1) the ability to provide true national benchmarks, and (2) the expertise that underlies the programs from the development of guidelines and performance measures, and the integration of those measures using a standardized, clinical lexicon that has been developed by NCDR. "Trying to generate clinical insights out of [claims data] occasionally works well. But most of the time," Masoudi says, "it obscures numerous clinical nuances that are highly relevant to the patients we take care of." Such nuanced data can yield valuable—and sometimes surprising—benefits to individual practices as well. "When you pay attention to the data…it helps improve your treatment and your documentation, and then you have a better quality score," says Nicholas A. Vaganos, MD, a cardiologist at Cardiology Consultants of Philadelphia, which participates in NCDR's PINNACLE Registry. His colleague, Christine Coyne, RN, director of quality and compliance, concurs: "Sometimes, as a large practice, we have this feeling that we are doing things very well…But then we get the numbers, and we realize, 'Oh, not quite as good in this particular area as we thought.'" Being in a registry is necessary, Vaganos says, "because there's no other way to keep track of all these parameters, measures, things like that. I honestly don't know how someone could do without this." Registry analytics also can help practitioners validate what they are doing right, Smith suggests, and better understand patient populations such as individuals with low back pain, Parkinson disease, stroke, or joint replacement. "If PTs are seeing great trends with great outcomes with a patient population," she says, "what are they doing differently from their colleagues, and how can their clinic and others adopt those interventions? How can they change the patient care pathway to achieve the best outcomes?" Identifying Flaws in Documentation Sometimes less-than-optimal outcomes data may reflect incomplete documentation rather than poor care, says Nathan Glusenkamp, newly appointed director of orthopedic registries at the American Academy of Orthopaedic Surgeons (AAOS). He had been president of provider solutions at FIGmd, the software vendor that powers APTA's registry and several others. Glusenkamp describes a client that was shocked that its performance measure for smoking cessation was at 0% even though the providers were sharing pamphlets and resources with every patient who was a smoker. As it turned out, smoking cessation was not being documented in the client's EHR system—a fact illuminated by its registry. Making that small change in documentation wasn't a huge burden, yet it improved providers' scores in the registry and, Glusenkamp says, had "other benefits for the patients, and for other providers who were seeing those patients," because other providers in the system could see that counseling already had been provided to a patient. Presenting a Complete Picture? Some PTs may have valid concerns about how their data might look to others, because they suspect the outcomes might be influenced by "social determinants of health" such as a patient's socioeconomic status or education level. While registries contacted by PT in Motion aren't yet measuring these factors, Coyne believes providers will "begin to see changes in what is being measured" under the Medicare Access and CHIP Reauthorization Act of 2015's [MACRA] Quality Payment Program (QPP), and "these kinds of things will begin to play into our [outcome] measures." The Physical Therapy Outcomes Registry intends to include risk adjustment in the calculation of outcome measures, Smith says. But, she continues, while "capturing the entire picture is incredibly important," the registry's first responsibility is helping PTs "understand their own practice before they even begin to compare themselves with similar practices." That, she says, will ensure that documentation and billing processes are "fine-tuned" to guarantee complete data. To help PTs dig deeper into their data, the APTA registry—which officially launched in February—will include "modules": condition- or disease-specific sets of data elements designed to describe and risk-adjust process-of-care and clinical outcomes for a defined patient population. The first such module, a collaboration with APTA's Academy of Pediatric Physical Therapy, will focus on congenital muscular torticollis. It will include neck movement and functional limitations, torticollis severity, and prognostic factors. Analyzing a large amount of such data will help describe typical practice patterns, variations in them, and associated outcomes. "Modules provide a continuous feedback loop not only for PTs, but also for guideline developers," says James Irrgang, PT, PhD, ATC, FAPTA, scientific director of the APTA registry's Scientific Advisory Panel. Meeting the Challenges Despite the promise of registries to improve quality of care and patient outcomes, there can be challenges when onboarding. Coyne emphasizes thinking twice "if you think you are going to join a registry and they are going to do everything for you." "I didn't have anyone to guide us, because [registries were] so new," she says. "But as somebody who's done this for several years, I would say the most important thing [as a registry user] is to keep in touch with your representative, question any data you think doesn't look right, and explore your data," Coyne says. "It's not just 1-sided." For example, she says, a registry will not gather all your data if you haven't directed them to where the specific data resides in your EHR. Providers always should check data output after an EHR update, as well, she adds, to make sure nothing has been lost in translation. EHR interoperability has been enough of a challenge to be a continuing priority of the Office of the National Coordinator for Health Information Technology (ONC). In the past, EHR vendors have been accused of "data-blocking"—hiding data, or keeping it from being shared with third-party software systems. This practice now is illegal under the 21st Century Cures Act. Rich, who recently served on an interoperability panel convened by the National Quality Forum, said the ONC likes the idea of "starting from the bottom up" and asking "What does your profession need to measure?" FIGmd—the registry software used by APTA, AAO, and ACC—has a very large clinical vocabulary. It uses machine learning and natural language processing to extract data—including narrative data—to the registry itself. (See "How Does the Physical Therapy Outcomes Registry Work?" on page 37.) Glusenkamp describes this artificial intelligence as "a matter of repetition." He explains, "When we work with a given specialty area for a length of time, a lot of that initial data mapping starts to be automated, and we can train our mapping software to do a lot of that matching up…as opposed to doing that manually. It reduces a lot of the burden on our team, and on the PT." As technology makes it easier to measure and monitor quality of care, those interviewed for this article say, there is less reason to shy away from registries. In the case of the first ACC registries, Masoudi describes 2 early responses to what he calls the "nascent movement to improve quality": to "stick one's head in the sand" or "accept one's professional responsibility and take a leadership position." He acknowledges it was a "tough road," but he adds, "as times have changed and the culture has changed…it's not enough just to say we're doctors and assume we deliver high-quality care. We must embrace accountability based upon data for the care we deliver." Michelle Vanderhoff is manager, editorial services, at APTA. How Does the Physical Therapy Outcomes Registry Work?If you are considering participating in the Physical Therapy Outcomes Registry, you might be asking, "Will we have to change our workflow?" and "Will it be a huge time commitment for staff?" The answer to both questions is "no." It doesn't require you to hire new staff, either.In the past, clinical registries relied on manual entry of patient data into an online portal, which—most providers agree—is impractical due to time commitment and staffing concerns. Interoperability between EHRs and registries has been challenging, too. FIGmd, the company behind the Physical Therapy Outcomes Registry, sidesteps the technological obstacles by using keywords, machine learning, and natural language processing to extract data. It includes narrative data and coded data such as billing codes, lab results, and prescriptions to give the most detailed picture of a patient episode. "The only potential change [to workflow] may come in the information providers gather on a patient so that we can capture as much clinical information as possible," says Kellin Lawler, client account manager at FIGmd. She cites 1 new client that had just implemented an EHR and needed guidance on adding certain data elements, such as ICF codes, that were not included in the EHR. After a practice completes the enrollment process, it works with a FIGmd representative to install "Registry Practice Connector" software either on the practice's EHR database server or on a computer that can access the database. This software acts as a go-between by querying the EHR, then securely pushing that data to the registry servers. For practices unable to install the software, the registry can accept data files from the practice, or its EHR vendor can push the data to the registry. During the integration process, the registry software will learn how to accurately "map" your data—essentially, where you keep the plates and the forks in your EHR's kitchen. Cheryl Dimapasoc, PT, DPT, director of implementation and compliance at OptimisPT, explains, "We take the measures and data in OptimisPT and match them up with the data that the registry is collecting. Once the names of those data points are mapped on both sides, there is a seamless flow of information from OptimisPT to the registry that allows that data to be analyzed."Nightly, the software uploads the encrypted data to the registry, separated into de-identified clinical data and public health information data, which is secured in compliance with Health Insurance Portability and Accountability Act laws and regulations. For practices that voluntarily participate in the Merit-based Incentive Payment System, that data will be re-identified before submission to the Centers for Medicare and Medicaid Services. After integration is complete, you may look at your registry dashboard, see a metric that appears inaccurate, and find that your documentation processes need to be fine-tuned, or that your mapping needs to be refined.Clients will be able to access site- and provider-level performance metrics to monitor quality measures and make adjustments where necessary. Practices will be able to track patients across multiple episodes of care and view aggregate outcomes data over time. Many of the provider reports will allow users to filter the data on demographics/characteristics in order to get a more accurate picture of their influence. Multisite practices can compare outcomes metrics and patient demographics among individual facilities.The registry takes security seriously. FIGmd maintains accreditation with the Data Registry Accreditation Program of the Electronic Healthcare Network Accreditation Commission (EHNAC), and is certified by the Office of the National Coordinator (ONC) for calculation and transmission of electronic clinical quality measures.